Characterization of Pathogens for Potential Diagnostic Tests
Characterization of Pathogens for
Potential Diagnostic Tests
A dissertation submitted to the
Graduate School
of the University of Cincinnati
in partial fulfillment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY (Ph.D.)
in the Department of Molecular Genetics, Biochemistry, & Microbiology
of the College of Medicine
by
KAREN M. GALLEGOS
B.S., Universidad Nacional Agraria La Molina, Perú
2013
Committee Chair: Alison A. Weiss, Ph.D. ABSTRACT
Traditional methods for detection of microbial pathogens, which are based on antibody recognition, often are limited by microbial antigenic changes. Characterization of pathogens and identification of stand-alone features with diagnostic value has the potential to replace the typical approaches of current diagnostic test. The approach taken here is to use invariant features of pathogens to develop new diagnostics and potential treatment. In this thesis, we characterized intrinsic features of two very common human pathogens: influenza and the main virulence factor of E. coli O157:H7, Shiga toxin.
In the first part of this study, we characterize influenza neuraminidase (NA) activity as a marker for influenza detection. Since influenza NA activity is an invariant characteristic of this virus, our objective was to develop an assay to detect and differentiate influenza NA activity from other respiratory pathogens such as parainfluenza and Streptococcus pneumoniae. We analyzed the influence of pH, calcium and NA inhibitors (oseltamivir and zanamivir) on NA using intact, viable pathogens. We found that these three pathogens display different pH optimum and the calcium requirement was different in each pathogen NA activity. We also found that influenza NA with H274Y mutation, which confers oseltamivir-resistance, displayed up to ten fold- difference in levels of drug resistance. Our results show that measuring NA activity in different conditions can be used to detect and differentiate NA activity of influenza virus, parainfluenza and S. pneumoniae.
In the second study, we analyzed the ability of host factors to bind influenza surface proteins as an alternative to replace antibodies as capturing agents. Three types of host factors were considered here: (1) natural (e.g. fetuin) and synthetic receptors containing sialic acid, which mimic the traditional influenza receptor; (2) host proteins binding to the invariable glycosylation pattern of influenza glycoproteins, high mannose core, (e.g. mannose binding
ii | P a g e lectin) and (3) factors playing a role in influenza-host interaction (e.g. galectin). Binding of influenza to host factors was assessed by a capturing assay. Our results show that influenza virus can be captured using all three approaches, and all of these factors show promise for replacing classical antibodies in diagnostic applications.
In the third study, we characterized the binding of two immunologically distinct forms of
Shiga toxin (Stx1 and Stx2) to understand their differences in cell toxicity and different affinity to the reported receptor, the glycolipid globotriaosylceramide (Gb3). Binding of Stx1 and Stx2 to
Gb3, various glycans, glycosphingolipids and their mixtures was studied in the presence or absence of membrane components, phosphatidylcholine and cholesterol. Our results show that in addition to the glycan, Stx2 binding is also influenced by the Gb3 lipid and the Gb3 lipid environment; while Stx1 binds primarily to the Gb3 glycan. These results show that Stx1 and
Stx2 recognize different residues in the Gb3 molecule and that Gb3 lipid environment modulates
Stx binding; this is especially true for Stx2. Our results contribute to the potential development of new therapeutics and diagnostic tests.
Overall, our results demonstrate that studying invariant characteristics of pathogens show promise to improve current diagnostic assays.
iii | P a g e
iv | P a g e
Dedicado a mis padres, mis hermanos
y a mi amor Rafael
por ellos, por siempre y para siempre
KMGV
Cada día sabemos más y entendemos menos
Albert Einstein
No des sólo lo superfluo, da vuestro corazón
Madre Teresa de Calcuta
v | P a g e
ACKNOWLEDGMENTS
En primer lugar, quiero agradecer a mi familia, mis padres Zulema y Alberto, a mis hermanos Miguel, Giancarlo, Francisco y Susan por siempre brindarme su constante apoyo, cariño, tranquilidad y por hacerme recordar que no hay cosas más importantes en la vida que la de tener una familia unida, llena de amor, risa y franqueza. Así también quiero agradecer a
Rafael Masitas por su soporte, su amor, su apoyo, su comprensión, su ayuda incondicional y por convencerme varias veces de no desistir y alcanzar mis metas.
I am grateful to Dr. Alison Weiss for giving me the opportunity to work my Ph. D. in the exciting field of influenza diagnostics and Shiga toxin. I benefited a lot from her enormous experience and scientific guidance, which allow my scientific mind to sharpen. I am sincerely thankful to her. I also would like to acknowledge the members of my dissertation committee —
Dr. A. Herr, Dr. R. Thompson, Dr. T. Thompson, and Dr. M. Staat— who provided helpful insight and feedback regarding this work.
My thesis could not have been completed without the help my friends in the program, who listen patiently while I tried to understand the wonderful world of influenza. Special thanks to my lab mate, Sayali Karve for giving valuable support to completed this thesis.
Finally, warm and heartfelt thanks to my family and my love Rafael, for making it all worthwhile and reminding me that the life outside science holds me many warm sunny days at
Peru’s best beaches with awesome food and the best company ever.
vi | P a g e
TABLE OF CONTENTS
ABSTRACT ...... 2
ACKNOWLEDGMENTS ...... 6
TABLE OF CONTENTS ...... 7
FIGURES AND TABLES ...... 12
ABBREVIATIONS ...... 15
CHAPTER I: BACKGROUND ...... 1
1 Introduction ...... 1
2 Biology of influenza virus ...... 2
2.1 History ...... 2
2.2 Classification ...... 2
2.3 Characteristics of influenza virus ...... 3
2.4 Influenza life cycle ...... 8
2.4.1 Binding to receptor ...... 9
2.4.2 Internalization into endosome...... 9
2.4.3 Fusion ...... 10
2.4.4 Uncoating...... 12
2.4.5 Import of RNP to the nucleus ...... 13
2.4.6 Replication of genome ...... 13
2.4.7 Nuclear export and assembly ...... 14
vii | P a g e
2.4.8 Virus budding and packaging ...... 15
2.4.9 Viral release ...... 16
2.5 Immune response to influenza infection ...... 17
2.6 Epidemic and pandemic influenza ...... 19
2.6.1 Antigenic Drift ...... 19
2.6.2 Antigenic Shift ...... 20
3 Developing diagnostic test ...... 21
3.1 Virus cell culture ...... 22
3.2 Serology ...... 22
3.2.1 Hemagglutination inhibition (HAI) ...... 22
3.3 Reverse-transcription polymerase chain reaction (RT-PCR) ...... 23
3.4 Neuraminidase activity ...... 24
3.5 Immunofluorescence ...... 24
3.6 Rapid Diagnostic Tests for influenza (RIDT) ...... 24
3.7 Pyrosequencing ...... 25
3.8 Need for a diagnostic test ...... 26
4 Major influenza antigens ...... 27
4.1 Hemagglutinin binding preferences ...... 27
4.1.1 Receptor binding site and binding preferences ...... 27
4.2 Neuraminidase ...... 31
4.2.1 Receptor binding site ...... 32
4.2.2 Calcium effect on influenza NA ...... 33
viii | P a g e
4.2.3 Neuraminidase inhibitors ...... 35
5 Capturing of viral particles ...... 41
5.1 High manose oligosaccharides ...... 41
5.2 Dendritic cell-specific ICAM grabbing non-integrin (DC-SIGN) ...... 42
5.3 Neuraminidases on microorganisms – Possible false positives ...... 45
CHAPTER 2: STUDIES ON NEURAMINIDASE ACTIVITY ...... 49
1 Abstract ...... 49
2 Introduction ...... 49
3 Experimental procedures ...... 53
3.1 Viruses, cells and recombinant NA ...... 53
3.2 Hemagglutinin (HA) titer ...... 57
3.3 Influenza Fluorescent Focus forming Assay (FFA) ...... 57
3.4 Real time Reverse Transcription-Polymerase Chain Reaction (RT-PCR) ...... 57
3.5 NA pH dependence ...... 58
3.6 NA calcium dependence ...... 59
3.7 Limit of detection of NA assay ...... 59
3.8 NAI susceptibility assay ...... 60
4 Results ...... 60
4.1 Characterization of the optimal pH for NA activity ...... 61
4.2 Characterization of the role of calcium in NA activity ...... 62
4.3 NAs from different phylogenetic sources display unique calcium and pH profiles. 64
4.4 Limit of detection of NA assay ...... 65
ix | P a g e
4.5 NAI susceptibility assay ...... 66
5 Discussion ...... 68
CHAPTER 3: STUDIES ON INFLUENZA BINDING ...... 73
1 Abstract ...... 73
2 Introduction ...... 74
3 Experimental procedures ...... 77
3.1 Viral preparations and capture molecules ...... 77
3.2 Capturing assay ...... 77
4 Results ...... 79
4.1 Influenza binding to synthetic sialylated receptors ...... 80
4.2 Influenza binding to protein host factors ...... 81
4.3 Influenza limit of detection using Fetuin and MBL with in a capturing assay...... 83
5 Discussion ...... 85
CHAPTER 4: SHIGA TOXIN BINDING ...... 88
1. Abstract ...... 88
2. Introduction ...... 89
3. Materials and Methods ...... 91
3.1. Production of recombinant Stx toxoids and B-pentamers ...... 91
3.2. Isothermal Titration Calorimetry (ITC) ...... 93
3.3. Glycan array studies ...... 93
3.4. Glycolipid ELISA ...... 94
3.5. Vero protection studies ...... 94
4. Results ...... 95
4.1. Characterization of individual glycan binding sites by ITC ...... 95
x | P a g e
4.2. Stx binding to synthetic glycans ...... 96
4.3. Stx binding to native Gb3 glycolipid ...... 97
4.4. Stx binding to glycolipid complexes ...... 98
4.5. Contribution of the ceramide to Stx binding ...... 102
4.6. Contribution of cholesterol to Stx binding ...... 104
4.7. Stx cellular toxicity in vero protection assay ...... 105
5. Discussion ...... 106
CHAPTER 5: FUTURE RESEARCH DIRECTIONS ...... 111
REFERENCES ...... 119
APPENDIX ...... 154
1. Effect of pH on Neuraminidase activity ...... 154
2. Sequence alignments of NA genes of H1N1pdm strains ...... 160
3. Sequence alignments of NA genes of human H1N1 strains ...... 161
4. Sequence alignments of NA genes of H3N2 strains ...... 162
5. Sequence alignments of NA genes of influenza type B strains ...... 163
6. Sequence alignments of H3 genes of influenza H3N2 strains ...... 164
xi | P a g e
FIGURES AND TABLES
TABLES
Table 1 1. Comparison of characteristics of influenza A, B, and C.
Table 1.2. Influenza A genes and proteins products and functions
Table 1.3. Influenza A and B mutations conferring resistance to NAIs from clinical isolates
Table 1.4. Possible false positive viral human pathogens
Table 2.1. Summary of characteristics of influenza strains and recombinant NA protein evaluated in Chapter 2
Table 3.1. Summary of characteristics of influenza strains evaluated in Chapter 3
Table 4.1. Plasmids and primers used in Chapter 4
Table 4.2. Glycolipids used in this study
FIGURES
Figure 1.1. Transmission electron micrograph images of Influenza virus
Figure 1.2. Model of Influenza A Virus
Figure 1.3. Structure of the influenza HA
Figure 1.4. Model of life cycle of influenza
Figure 1.5. Conformational changes of HA during membrane fusion
Figure 1.6. Structure of the influenza M2 bound to adamantane
Figure 1.7. Schema of occurrence of influenza pandemics and epidemics in relation to level of immunity in the population
xii | P a g e
Figure 1.8. Example of Structure poly antennary glycan chains found on fetuin
Figure 1.9. Structure of the influenza N2 bound to Sialic Acid and Ca2+
Figure 1.10. Structure of NA substrate and inhibitors
Figure 1.11. Mechanism of Action of Neuraminidase Inhibitors
Figure 1.12. Mechanism of resistance to oseltamivir
Figure 1.13. Glycosylation pathway
Figure 1.14. DC-SIGN structure model
Figure 2.1. Diagrams of NA sequence and structure
Figure 2.2. Influence of pH on NA activity
Figure 2.3. Influence calcium on NA activity
Figure 2.4. Comparison of NA from different phylogenetic groups
Figure 2.5. Limit of detection of influenza and S. pneumoniae NA activity
Figure 2.6. Evaluation of co- and pre- incubation of oseltamivir on NA activity
Figure 3.1. The N-glycosylation process
Figure 3.2. Glycan structures
Figure 3.3. Evaluation of binding of synthetic sialylated glycans to influenza virus
Figure 3.4. Evaluation of the recombinant host proteins to capture influenza virus
Figure 3.5. Evaluation of the limit of detection of Fetuin as a capturing molecule for influenza virus
Figure 3.6. Evaluation of the limit of detection of MBL as a capturing molecule for influenza virus
Figure 4.1. Binding of Stx1 and Stx2 to purified Pk trisaccharide and P tetrasaccharide by
ITC
xiii | P a g e
Figure 4.2. Glycan array results for Stx1
Figure 4.3. Stx binding to pure Gb3
Figure 4.4. Binding of Stx1 and Stx2 to purified glycolipids and mixtures
Figure 4.5. Stx binding to Gb3, Gb4 and Gb3/Gb4 mixture in the presence of cholesterol and phosphatidyl-choline
Figure 4.6. Stx binding to Gb3 analogs
Figure 4.7. Stx binding to Lyso-Gb3
Figure 4.8. Comparison of Stx binding to Gb3 in absence of cholesterol or phosphatidylcholine
Figure 4.9. Stx binding to Gb3 in presence of a cholesterol analog
Figure 4.10. Vero protection studies
Figure 5.1. Detection of influenza in patient samples by NA activity assay
Figure 5.2. Evaluation of X-NeuAc substrate
Figure 5.3. Evaluation of p-NP-NeuAc substrate
Figure 5.4. Evaluation of different reaction times and temperatures
Figure 5.5. Relative NA activity of influenza strains comparing different reaction times, incubation temperature and final pH
Figure 6.1. Electrostatic potential map of neuraminidases
Figure 6.2. Scheme of the mechanisms of influenza NA
xiv | P a g e
ABBREVIATIONS
α alpha
β beta
δ delta
η nano
λ lambda
μ micro
°C degree Centigrade
Å angstrom
A, Ala alanine
AB5 bacterial toxin with one active subunit and five binding subunits
ABS absorbance
C, Cys cysteine
Ca2+ Calcium ion
CAPS 3-(cyclohexylamino)-1-propanesulfonic acid
CFG Consortium for Functional Glycomics
CHES N-cyclohexyl-2-aminoethanesulfonic acid
Ct cycle threshold
C-terminus carboxyl terminus
D, Asp asparagine
DC-SIGN Dendritic Cell Specific ICAM3 Grabbing Nonintegrin dH2O distilled water
DNA deoxyribonucleic acid
xv | P a g e
E, Glu glutamic acid
EDTA ethylenediaminetetraacetic acid
EGFR epidermal growth factor receptor
ELISA enzyme-linked immunosorbent assay
ER endoplasmic reticulum
ER endoplasmic reticulum
F, Phe phenylalanine
FFA fluorescent focus forming assay
FFU fluorescent focus forming units
G, Gly glycine
Gal galactose
GalNAc N-acetylgalactosamine
Glc glucose
GlcNAc N-acetylglucosamine
H hour
H, His histidine
HA hemagglutinin
HCl hydrochloric acid
HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
HN hemagglutinin-neuraminidase
HUS hemolytic uremic syndrome
I, Ile isoleucine
IFN interferon
IgG immunoglobulin G
IL interleukin
xvi | P a g e
IPTG isopropyl β-D-thiogalactopyranoside
K kilo
K, Lys lysine
Kd dissociation constant kDa kilodalton
Ki inhibition constant
L liter
L, Leu leucine
LB Luria-Bertani
LD50 50% lethal dose
LPS lipopolysaccharide m meter
M Molar
M, Met methionine
Man mannose
MBL mannose binding lectin
MDCK cells Madin-Darby canine kidney cells min minute
MMR macrophage mannose receptor
MW molecular weight n nano
N Nitrogen
N, Asn asparagine
NA neuraminidase
NaCl sodium chloride
xvii | P a g e
NAI neuraminidase inhibitor
NaOH Sodium hydroxide
NeuAc N-acetylneuraminic acid nH Hill slope
NIH National Institutes of Health
NMR nuclear magnetic resonance
N-terminal amino terminal
OD optical density
P, Pro proline
PAMP pathogen-associated molecular pattern
PBS phosphate buffered saline
PCR polymerase chain reaction
PDB Protein Data Bank
PEG polyethylene glycol
Q, Gln glutamine
R, Arg arginine
RBC red blood cells
RFU relative fluorescent units
RNA ribonucleic acid
RU response units
S, Ser serine
SD standard deviation
SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis
SPR surface plasmon resonance
T, Thr threonin
xviii | P a g e
Tris tris(hydroxymethyl)aminomethane
V, Val valine
W, Trp tryptophan
WHO World Health Organization
WT wild-type
Y, Tyr Tyrosine
xix | P a g e
CHAPTER I: BACKGROUND
1 Introduction
Influenza virus is a highly contagious agent that causes upper and lower respiratory tract infection due to its airborne nature and ability to infect cells lining the airways of the host.
Influenza can produce local outbreaks or epidemics resulting in up to 500,000 deaths per year worldwide [1]. In United States, it is responsible for 200,000 hospitalizations and 36,000 deaths per year [2]. Influenza illness is an acute and frequently self-limited infection that can be caused by different types of influenza virus (e.g. A or B) occurring in seasonal, pandemic, and zoonotic outbreaks. The most common symptoms of this illness are fever, malaise, and cough [3, 4].
Two of the most important antigenic proteins of influenza, hemagglutinin and neuraminidase, naturally accumulate point mutations (antigenic drift), helping the virus escape from host antibody recognition. Moreover, influenza segmented genome allows the virus to exchange new genes from a distantly related virus (antigenic shift), resulting in the emergence of new pandemic influenza strains. Current diagnostic tests for influenza (e.g. immunofluorescence assays, serology, viral culture and PCR) encounter antigenic drift variation and require for expert laboratory skills. A rapid diagnostic influenza test, which is quick and reliable, could significantly reduce the disease complications by identifying infected individuals and preventing transmission. Additionally, the development of a rapid diagnostic test for the identification of antiviral-resistant influenza strains will allow an efficient use of antivirals drugs.
The objective of this study is to develop a new approach for the diagnosis of influenza based on invariable aspects of the influenza biology, which are not subject of antigenic shift or drift. We will use the glycosylation pattern of influenza glycoproteins and the neuraminidase activity required for the progression virus infection. In this thesis, we tested capturing molecules to efficiently capture influenza and used the enzymatic activity of the influenza neuraminidase for detection.
1 | P a g e
2 Biology of influenza virus
2.1 History
The origin of the influenza virus name was derived from the Latin word influentia, which means influence. In the 14th century, Italians believed that flu symptoms and disease were due to the influence of the stars [5]. Evidence of the influenza disease can be found as far back as
2,500 years ago, when Hippocrates described the flu symptoms in 4th century BC. However, the causative agent of influenza was thought to be bacterial until 1933 when Smith et al. proved that bacteria-free nasal washings from influenza-type A patients could transmit the disease to ferrets suggesting that flu infection was not cause by a bacterium [4]. Later, in 1939 Francis was the first to isolate influenza virus (type B strain) followed in 1950 by Taylor’s isolation of influenza C.
In 1936, Burnet found that influenza was able to grow in embryonated hen eggs, thus allowing for the development of inactivated influenza vaccines [6]. Later, the discovery by Hint in 1941 of the ability of influenza to cause hemagglutination allowed for the expansion of inexpensive methods to measure virus content and specific antibodies reactivity [7]. The first live vaccine was approved 2003. In addition, there are two classes of anti-viral drugs approved for influenza treatment. The first type inhibits M2 ion channel of influenza A and includes amantadine and rimantadine. The second type inhibits neuraminidase activity and includes zanamivir and oseltamivir.
2.2 Classification
Influenza viruses, members of the family Orthomyxoviridae, are classified into three types, termed influenza A, influenza B, and influenza C virus. This classification is based on antigenic differences. The three major virus antigens (nucleoprotein, hemagglutinin and neuraminidase) are used in classification [5]. Since the nucleoprotein is a stable antigen, it is used for differentiation of influenza virus types. The nucleoprotein antigens of influenza A, B or
C do not exhibit any serological cross-reactivity unlike other major antigens, such as
2 | P a g e
hemagglutinin or neuraminidase, both of which are found on the surface of the viral membrane and serve as the major targets for antibodies immunizing against influenza [8].
Beside the differences in the nucleoprotein antigen, distinct influenza types also show main variations in genetic organization, structure, host range, epidemiology, and clinical characteristics. Some of these differences are summarized in Table 1.1.
All influenza types share major characteristics: the viral membrane, derived from the infected host cell); the surface glycoproteins hemagglutinin and neuraminidase; and segmented negative-sense single strand RNA as genetic material. The nomenclature of influenza virus is standardized and includes: the influenza type, geographic location of first isolation, strain designation, and isolation year. (e.g. B/Brisbane/60/2008 or A/Florida/03/2009) [3].
2.3 Characteristics of influenza virus
Influenza is an enveloped virus that can be filamentous (20 nm in diameter and 200-
3000 nm long) or spherical (80 to 120 nm in diameter) (Figure 1.1). It has been suggested that the morphology of the virions is associated with how the virus spreads. Filamentous virions are thought to be more efficient in infecting target adjacent cells where spherical viral particles are more suitable for aerosol spread between hosts. Influenza clinical isolates are usually filamentous virions; whereas egg- or cell-culture virions have only spherical morphology [9].
Table 1.1. Comparison of characteristics of influenza A, B, and C. Influenza A Influenza B Influenza C Genetics 8 gene segments 8 gene segments 7 gene segments 11 viral proteins 10 viral proteins 9 viral proteins Structure NB ion channel membrane M2 ion channel unique HEF* unique protein unique Humans, swine, equine, Host range avian, marine mammals, Humans only Humans and swine dogs, and cats Antigenic drift only. Two Antigenic drift only; Epidemiology Antigenic shift and drift main lineages co-circulate multiple variants (Yamagata and Victoria ) Causes large pandemics Severe disease generally in Clinical Mild disease; with high mortality in very elderly or high-risk people; features seasonality not seen young people and elderly pandemics not seen * HEF: hemagglutinin, esterase, and fusion glycoprotein
3 | P a g e
Figure 1.1. Transmission electron micrograph images of Influenza virus. (A) Colorized micrograph showing the spherical morphology of influenza H1N1pdm flu virus. Image ID 11214 from Public Health Image Library (PHIL) at CDC/ C Goldsmith and A Balish (2009). (B) Micrograph showing the filamentous morphology of influenza-A H1N1, (zoom 189,000x). 7814. PHIL at CDC (1978). (C) Colorized micrograph of Avian influenza A H5N1 viruses (filamentous and spherical gold A B C particles) grown in mammalian cells (green). Image ID 1841 from PHIL at CDC/ C Goldsmith; J Katz; S Zaki (1997)
Influenza viruses all have a segmented genome consisting of 7 or 8 linear negative-sense single-stranded RNAs [4]. The influenza genome is 12,000-15,000 nucleotides (nt) in total, with
RNA segments having sizes ranging from 2,300-2,500 nt to 800-900 nt. Whole genome sequences of many influenza viruses are available at www.fludb.org. The 12-13 nt comprising the ends of each viral RNA segment are important for gene transcription and are highly conserved among all eight RNA segments. The majority of influenza RNA segments encode for only one protein, and influenza genomic segments encode for all proteins found in the virion.
Reassortment of influenza genome is facilitated by the nature of its segmented genome [4].
All influenza types (influenza A, B and C) share seven viral proteins: hemagglutinin (HA), neuraminidase (NA), ribonucleoprotein (RNP), M1 matrix protein, and three proteins that form
RNA polymerase [4]. In addition to those proteins, influenza A has the M2 ion channel and influenza B has the NB membrane protein, a putative proton channel thought to play a role in virus entry into the host cell. Influenza C encodes a unique Hemagglutinin-esterase-fusion glycoprotein (HEF). Table 1.2 shows the genes and proteins products on influenza A virus.
The viral membrane surface is covered with spikes that vary from 10 to14 nm in length.
These spikes represent of the two major viral antigens, HA and NA (Figure 1.2). NA is found in clusters on the membrane, while HA is equally distributed in the surface. HA represents about
25% of total viral protein, while NA is only about 5% thus making a HA/NA ratio of 5:1 [8].
4 | P a g e
Table 1.2. Influenza A genes and proteins products and functions Segment Gene Name of the Proposed function N⁰ product protein Basic polymerase 1 PB RNA transcriptase 1 Basic polymerase Cap-binding; endonucleolytic 2 PB2 2 Cleavage Stabilizes polymerase complex, endonuclease 3 PA Acidic polymerase activity, cap- and promoter binding Viral attachment to host cell; fusion and 4 HA Hemagglutinin internalization Cleaves sialic acid, release from host cells; 5 NA Neuraminidase prevents aggregation Encapsidates RNA; regulation of transcription 6 NP Nucleoprotein and replication 7 M1 Matrix 1 Forms viral core; controls RNA nuclear export M2 Matrix 2 Ion channel; needed for uncoating Antagonizes interferon and involved in mRNA 8 NS1 Non-structural 1 transport from nucleus Nuclear export NEP* Transports new RNP from nucleus to cytoplasm protein * Formerly called NS2
All influenza HAs are glycosylated proteins with a single-pass through the membrane
(considered a type I membrane protein). HA is the influenza virus receptor-binding and membrane fusion glycoprotein; it also is the target for host immune response such as infectivity- neutralizing antibodies. HAs attach to receptors on the cell surface that contain sialic acid
(NeuAc). This protein is also responsible for viral penetration, internalization and fusion with the endosomal membrane. These processes will be discussed further in Chapter 1 section 2.4. The specificity of HA to certain conformations of NeuAc determines the host-range restriction and virulence of a specific influenza strain [8]. HA binding preferences and its importance for the development of a new diagnostic test will be discussed in detail in Chapter 1 section 4.1. Our results in influenza binding to natural and synthetic sialylated receptors are discussed in
Chapter 3.
5 | P a g e
Figure 1.2. Model of Influenza A Virus. NA and HA are shown in light red and blue, respectively. M2 ion channel and RNA nucleoprotein are shown in light purple and light green respectively. The inner pink lining represents the M1 matrix protein. Modified from Free resources at Centers for Disease Control and Prevention, National Center for Immunization and Respiratory Diseases (NCIRD)
Functional HA has a molecular weight of 225 kD and is formed by three identical monomers (75 kD each) [10]. Each HA monomer is formed by polypeptides HA1 (50 kD) and
HA2 (25 kD). HA1 and HA2 are kept together by disulfide bonds. HA is synthesized as a precursor, HA0, which needs to be cleaved to become fully active. HA trimerizes in presence of chaperones in the endoplasmic reticulum (ER), where it is also glycosylated. The viral protein is then palmitoylated at the pre-Golgi site and escapes from the ER before moving to the Golgi network where mannose residues are trimmed [11]. A signal present in HA transmembrane domain targets HA to the apical plasma membrane during viral replication in epithelial cells. On the cell surface, it is associated with detergent-resistant lipid rafts that contain high levels of glycosphingolipid and cholesterol [12]. Inactive HA is cleaved into HA1 and HA2 outside of the host cell by one or more endoproteases that are secreted by the bronchial epithelial cells in the lung including furin protease (found in Golgi apparatus and host cell surface), respiratory-tracts extracellular proteases [9] or type II transmembrane serine proteases [13]. HA is also responsible the viral penetration, internalization and fusion with the endosomal membrane.
Like influenza A and B HA, influenza C HEF is responsible for binding NeuAc on the cell surface, determining the host range restriction, and facilitating viral penetration and
6 | P a g e
internalization. Unlike HA, HEF is able to cleave NeuAc off of any receptor on the host cell surface. This activity is important during virion release. In influenza type A and B, NA is responsible for this sialidase activity.
As mentioned above, NA catalyzes the cleavage of terminal NeuAc residues – α2,3, or
α2,6 – from viral and host cellular glycoproteins and glycolipids. During virus budding, NA cleaves NeuAc from the glycosylated HA and host proteins to facilitate virus release [7]. NA activity also prevents self-aggregation, guaranteeing the efficient spread of progeny virus from host cell to host cell. If NA activity is inhibited, infection is limited to one round of replication and viral particles are prevented from invading the upper airway. Recently, it has been shown that
NA activity in late endosomes (acidic pH) increases viral replication by using influenza strains with low-pH-stable NA versus influenza strains with low-pH-unstable NA. [14]. Some antivirals such as oseltamivir and zanamivir inhibit NA activity [15]. The emergence of influenza strains resistant to neuraminidase-inhibitors will be discussed in detail in Chapter 1 section 4.2. HA and NA genes plays a main role in the viral genetic variation by single point mutations or reassortment of segments resulting in the appearance of new influenza strains [16].
The third protein shared by influenza viruses is the RNP, which is comprised of viral RNA encapsidated by nucleoprotein (NP). NP binds each RNA segment to form the helical RNP structures found in influenza virions. The main function of NP is to encapsudate the viral genetic material to facilitate RNA transcription, replication and packaging [17]. NP is reported to also bind cellular proteins including actin, nuclear transport proteins and nuclear RNA helicases [18].
It has been suggested that NP participates in the viral transcription-replication processes, coating the newly cRNA or modulating the viral polymerase activity via binding to them [19].
All influenza virus encodes a RNA polymerase that catalyses both viral RNA transcription
(viral RNA → mRNA) and replication (viral RNA → cRNA → viral RNA) of the viral genome during infection of host. Influenza RNA polymerase is a heterotrimer formed by three subunits:
7 | P a g e
PB1, PB2, and PA; each one is encoded in different RNA segments. While PB1 has the polymerase activity, all three subunits are required for RNA replication cells [20, 21].
Two different non-structural proteins, NEP and NS1, are encoded by different reading frames of NS gene in the influenza genome. These viral proteins are found in the host cytoplasm but not within the virion itself. The nuclear export protein (NEP) mediates the export of viral RNPs [22]. The non-structural (NS1) protein is a non-essential virulence factor that serves multiple functions during infection. The major role of NS1 is to inhibit host immune responses by preventing both interferon (IFN) production and downstream antiviral effects of
IFN-induced proteins [23]. Other roles have also been attributed to NS1, including the regulation in viral RNA synthesis, viral morphogenesis, viral mRNA splicing [24], enhancement of viral mRNA translation, and suppression of host apoptotic responses [25].
Influenza M gene also encodes for two proteins using different reading frames of the same
RNA segment: M1 matrix protein and M2 ion channel. M1, the most abundant protein in the virion, plays critical roles during steps throughout the viral replication cycle, from virus entry to assembly and budding of virions [26]. The main role of M1 is to form an inner core underneath the viral membrane, which binds RNP [27]. If M1 is expressed alone without expression of other viral proteins, it can form virus-like particles in transfected cells [28]. M1 is also responsible for the shape of the virions (e.g. spherical or filamentous). The second protein encoded by M gene is the M2 ion channel, a homotetramer. M2 is found on the envelope surface of influenza A strains only and is present in low quantities (only 16 - 20 molecules per virion). M2 is a proton- selective ion channel that is essential for the efficient release of the viral genome during virus entry [29].
2.4 Influenza life cycle
Replication of influenza virus is very rapid, taking only 6 hours for the first virion to exit from infected cells. The influenza life cycle is divided into different stages: host binding,
8 | P a g e
internalization, fusion, nuclear import of RNP, transcription, replication, and viral protein synthesis, nuclear export of RNP and virus assembly and release [30].
2.4.1 Binding to receptor
The influenza virus HA binds to host cell sialylated receptors. The receptor-binding site of the viral HA will be discussed further in Chapter 1 section 4.1. Figure 1.3 shows the crystal structure of the functional mature hemagglutinin and NeuAc binding sites, which are located in the globular domain of the molecule. Figure 1.4 shows that the binding is the first step in virus life cycle.
Figure 1.3. Structure of the influenza HA. The HA monomer (A) and trimer (B). In the trimer, HA2 is shown in light purple, orange and blue. HA1 is shown in yellow, green and sky blue. The fusion peptide is shown in red. The image was generated using Pymol and Protein Data Bank, code 3HMG.
2.4.2 Internalization into endosome.
Virion internalization is induced by HA (Figure 1.4, step 2). Bound viral particles are internalized into host cells. Many different host pathways have been shown to promote influenza internalization. Clathrin-dependent endocytosis (two-thirds of virus particles), caveolae- mediated, clathrin- and caveolin-independent pathway (one third of the virus particles), and/or macropinocytosis have all been implicated in influenza entry into host cells, but the exact mechanism of viral entry is still unclear [31]. Recent data suggest that spherical particles may activate receptor tyrosine kinases (e.g. epidermal growth factor receptor, EGFR), which trigger
9 | P a g e
intracellular signals that leads to formation of clathrin-coated pits and virus uptake [32].
Filamentous virions may use a different mechanism since their long length prevents them from fitting in the clathrin-coated pits. Other filamentous virus such as Ebola virus enters cells via macropinocytosis [33]. It is possible that filamentous influenza virions may also use macropinocytosis for internalization.
Figure 1.4. Model of life cycle of influenza. Numbers indicate different stages of the life cycle. Abbreviations used: cRNA (+), (+) sense complementary RNA; mRNA (+), (+) sense messenger RNA; NP, nucleoprotein; NS1, a non-structural protein, NS2, Pol, viral polymerases; vRNA (-), (-) sense genomic RNA; viral ribonucleoproteins, viral RNPs. Figure generated using Chemdraw.
2.4.3 Fusion
Since influenza is an enveloped virus with a bilayer lipid membrane as a part of its structure, the fusion of viral and host-cell membranes is required to liberate its internal genetic material (Figure 1.4, step 3). This fusion process is facilitated by HA. Crystal structures of HA in
10 | P a g e
the pre-fusion conformation, the post-fusion conformation, the uncleaved and non-functional precursor conformation [10] have contributed to the understanding the fusion of viral and host membrane mediated by HA. At the first step, HA1, NeuAc–binding domain, settled at the top of a stalk formed by HA2 (Figure 1.5A) forming the pre-fusion conformation [34]. The fusion peptide is kept inside the pre-fusion structure isolated by the N terminus of HA2 (Figure 1.3B, shown in red) [35]. When pH decreases in the endosome by the action of host ions channels,
HA undergoes conformational changes (Figure 1.5B). HA1 separates from HA2, but they remain covalently linked by disulfide bonds. The variable loops near the HA2 main helices then fold into helices, creating a long coiled-coil at the central part of the HA trimer causing the translocation of the fusion peptide toward the endosomal membrane (Figure 1.5C) [36]. The last step sees the lower parts of HA protein fold back, causing contact between fusion peptides and
HA transmembrane segments (Figure 1.5D) resulting in the post-fusion conformation. [34]. The folding is completed when the viral and host membranes have fused completely. NMR studies suggest that these fusion peptides may be lying partly immersed in the resulting bilayer membrane [36].
Figure 1.5. Conformational changes of HA during membrane fusion. Each subunit is shown in a different color Red asterisk represents the sequestered fusion peptide of the red subunit at the N terminus of HA2. (A) pre-fusion conformation. (B) Acid pH conformation. (C) Fusion con- formation. (D) post-fusion conforma-tion. Modified from Harrison S (2008) [34]
D
11 | P a g e
A B C
2.4.4 Uncoating.
Viral M2 ion channel plays an important role during this step in viral replication (Figure
1.4, step 3). Virus un-coating starts with endosome acidification, which triggers fusion, as well as, M2 activity. During fusion, HA will form a pore through which viral RNP can be released into the cytoplasm [37]. However, RNP is bound to inner core formed by M1 structural protein.
Before fusion is finished, M2 ion channel activity decreases the internal pH in the virion, disrupting the association of RNP, M1, and viral membrane [38]. Once the dissociation is complete, M1-free RNPs are liberated in the cytoplasm for import to the nucleus. Therefore, M2 enables RNA migration to the cell nucleus, where virus RNA transcription and replication occurs.
Some antivirals such as amantadine (1-adamantamine hydrochloride) and its analogue, rimantadine (α-methyl-1-adamantanemethylamine) target and inhibit M2 activity. These antivirals plug the ion channel [39]. Adamantanes do not affect influenza type B due to the lack of M2 and the fact that NB ion channel of influenza B has no sequence similarities with M2.
Occurrence of amantadine-resistant influenza strains is frequently rapid; hence, this drug is no longer used for human treatment. In 30% of patients treated with adamantanes, viral resistance develops within three days of treatment [40]. Moreover, the use of these drugs is associated with several central nervous system side effects (e.g. anxiety, depression, and insomnia) and gastrointestinal disorders (e.g. nausea, vomiting, and dyspepsia) [41].
Genotypic studies demonstrated that a single mutation (Ser31Asn) in the M2 gene confers antiviral resistance (Figure 1.6). Adamantane-resistant strains are genetically stable, transmissible and as pathogenic as wild type isolates [42]. The majority of current circulating influenza A strains are adamantane-resistant. Additionally to Ser31Asn mutation recent isolates have been reported to have other mutations [43]. Seasonal H1N1 viruses have specific M2 mutations (Val27Ala); while H3N2 isolates have the Ser31Asn, substitution [43]. Interestingly, pandemic H1N1 viruses do not contain the Ser31Asn mutation in their M2 gene, since they are believe to be originated from swine H1N1 strain; however, recent reports informs the occurrence
12 | P a g e
of adamantine-resistant pandemic H1N1 strains [44, 45]. Several new adamantine analogs and other more distant related compounds have been tested in vitro against influenza A; however, these compounds have not been studied in animals and humans [46].
Figure 1.6. Structure of the influenza M2 (PDB 2C9J) bound to adamantane. Three of the four identical chains of M2 ion channel are shown in green. Substrate binding residues are highlighted in red. Adamantine is shown in orange and blue. Figure generated using PyMol.
2.4.5 Import of RNP to the nucleus
Cytosolic RNPs are transported into the host nucleus after the recognition of their nuclear localization signal (NLS) [17]. Karyopherin beta nuclear transporter facilitates the transport of RNP. First, karyopherin beta forms a trimeric complex. Once the complex has formed, it is able to bind to the nuclear pore and the nuclear membrane (Figure 1.4, step 4).
The RNP are then transported and released into the nucleus [47].
2.4.6 Replication of genome
Once in the nucleus viral RNAs are transcribed and replicated (Figure 1.4, step 5). The viral (-) stranded RNA (vRNA) functions as template for synthesis of the capped and polyadenylated mRNA, and the full-length (+) stranded RNA [complementary RNA, cRNA]. The cRNA serves as a template for the synthesis of new viral RNA [48]. RNA polymerase enters to the nucleus with RNP complex and it catalyzes all of these three reactions (vRNA→ mRNA; vRNA→ cRNA→ vRNA) [48]. Unlike DNA polymerases with one error / 109 bases, influenza
RNA polymerase lacks proofreading activity and usually has one error / 104 bases. Therefore,
13 | P a g e
each round of replication results in mixed populations; some of which might be not viable, while others might have potentially advantageous mutations [49].
Influenza viruses need host cell nuclear functions to replicate including mRNA splicing and cap acquisition. During initial steps of replication, viral mRNAs are capped and polyadenylated. Interestingly, their caps are stolen from host cell mRNAs. These primers are formed by the cleavage of a subset of host cell RNAs by the endonuclease activity of PB2 [30].
Viral mRNA synthesis is initiated with 10 to 15-nucleotide long capped primers. Later, cap- primers are elongated by the RNA polymerase. The phosphoprotein PA is involved in stabilization of polymerase complex, endonuclease activity, cap- and promoter binding. Viral polyadenylation of mRNA takes place after the virus has shut down the host cell polyadenylation processing and has inhibited nuclear export of host pre-mRNAs [50, 51]. Viral poly-A tails are produced by the viral polymerase when it stutters on poly-U regions near the 5' end of the vRNA template. After viral RNA replication, newly synthetized M1 alone binds to ribonucleoprotein
(RNPs) in nucleus, inhibiting viral transcription. However, when Nuclear Export Protein (NEP) is associated with M1 and RNP, the whole complex is exported out of the nucleus. Interestingly,
M1 is not able to bind NP without viral genomic RNA [22, 27]. The exact mechanism by which any of these mRNAs are transported outside the nucleus for transcription is still uncertain.
However, recently, a part of the transcription-export complex, UAP56, is reported to play a role in inhibited the trafficking and/or the translation of the viral M1, M2 and NS1 mRNA [52].
2.4.7 Nuclear export and assembly
After the vRNA transcription, the assembly and nuclear export of RNP may then take place. vRNA is packaged into RNP complexes, which contain RNA, NP and M1 (Figure 1.4, step 10) [30]. NP nucleoprotein covers and protects the sugar phosphate backbone of the vRNA. M1 matrix protein is responsible for export of the whole complex from the nucleus. M1 mediates the interaction of the complex with the nuclear-export machinery. Recent data demonstrates that the lack of M1 SUMOylation might prevent the nuclear export of RNP for later
14 | P a g e
viral assembly [53]. The RNP complex does not interact directly with nuclear-export machinery.
The viral protein, NEP binds to the RNP in a M1-mediated reaction. Previous data demonstrate that the C-terminal domain of NEP (residues 81–100) is required for M1 binding and the nuclear export of new vRNPs [22]. This whole complex (RNP, NEP, and M1) interacts with nuclear- export machinery that allows it to cross the nuclear pore to the cytosol [12].
Previous studies show that NEP induces the formation of an export complex in which
NEP N-terminal associates with cellular export protein Crm1 and NEP C-terminal associates with M1 protein, which is already bound to viral RNPs (Figure 1.4, step 10) [22]. Unlike, influenza A NEP, influenza B NEP does not binds to M1 and associates directly to viral RNPs promoting its export independently of M1. Influenza NEP protein also binds several nucleoporins and is suggested to recruit export machinery and direct export of the whole complex.
2.4.8 Virus budding and packaging
Once HA, NA and M2 proteins are synthesized, they go into the endoplasmic reticulum to get properly folded, polymerized, and glycosylated. Later, they are exported to the Golgi apparatus where palmitoylation of cysteine residues occurs on HA and M2 [11]. In addition, M2 plays a role during the viral proteins intracellular transport. During assembly of new virions, M2 increases the internal pH of normally acidic Golgi network in such a way that M2 prevents the conformation change of new HA into the fusion-active conformation [38].
The budding site is initiated when HA, NA, and NP are associated into lipid rafts on the membrane causing the deformation of the bilayer (Figure 1.4, step 10). HA transmembrane
(TM) domain, together with palmitoylated cysteine residues on TM domain and cytoplasmic tail, mediates lipid raft association. Influenza requires the transport of viral proteins to the budding site for virus assembly [12]. The budding process itself requires three steps: bud initiation, bud growth, and bud completion, releasing the virus from host cells. Human influenza viruses bud from the apical surface of polarized epithelium into lung lumen. The exact mechanism of virus bud initiation is still unknown. However, HA and NA are the only viral proteins capable of
15 | P a g e
initiating the budding event. The budding starts by the association of M1 matrix protein with the cytoplasmic tail of NA and HA, which serve as docking sites for M1 [9]. The accumulation of M1 matrix protein at the cytoplasmic face of cell membrane leads to the development of the inner core and causes membrane bending and bud initiation [12]. M1 also works as a docking site for viral RNPs and recruits M2 to viral budding site. During last step of viral budding, M2 alters membrane curvature causing membrane scission and the release of the new virion (Figure 1.4, step 12) [54].
The mechanism of packaging of RNP containing vRNA genome is suggested to occur either randomly or selectively. The random packaging hypothesis is supported by the ratio of infectious particles to noninfectious particles obtained in viral stocks (1 over 400 particles assembled) which agrees with the probability of random packaging of 8 RNA segments. The selective packaging mechanism proposed that each RNA segment has a signal that allows it to be incorporated into virus particles. This is supported by the electron microscopy images that show the viral RNPs organized in a distinct pattern [55].
2.4.9 Viral release
Once the inner core is formed, the final budding stage starts and the virion needs to be released. Many viruses use the host Endosomal Complex Required for Transport (ESCRT) to mediate budding and scission; however, influenza is reported to use an ESCRT-independent mechanism [56]. Recently, M2 protein was found to play a role in modifying membrane curvature in a cholesterol-dependent manner. Moreover, during budding, M2 mediates membrane scission, and allows the virion to be released [56]. After membrane scission, the virion may still be attached to the cell membrane by HA binding to cell-surface NeuAc residues.
NA cleaves NeuAc off the cell surface, preventing the HA–receptor interaction and enabling the budded virion to be released.
Little is known about the involvement of host cellular factors, and much effort has been made to understand and identify all host cellular functions needed during influenza life cycle [57,
16 | P a g e
58]. A recent analysis of current data in this topic has reported that about 6% of human genes that encode proteins may be involved in influenza life cycle [57]. Compiling the data obtained in six independent siRNA screens of total host genome and narrowing down those identified genes by pair-wise comparison, a total of 128 human genes were found in at least two screens. Using bioinformatics tools, those 128 host genes were grouped in cellular functions such as endocytosis, translation initiation, and nuclear transport [57]. More detailed analyses of the function of these identified human genes will allow a better understanding of exact cellular host machinery activated during influenza infection.
2.5 Immune response to influenza infection
The innate immune responses to influenza are triggered by recognition of pathogen- associated molecular patterns (PAMPs) in the virus. Intracellular and extracellular pattern recognition receptors (PRRs) in host cell are activated during viral infection and initiate the innate immune responses. There are three families of PRRs: Toll-Like Receptors (TLRs),
Retinoic acid-Inducible Gene 1 protein (RIG-I)-like helicases (RLRs), and Nucleotide-binding
Oligomerization Domain (NOD) Leucine – rich - Repeat-containing proteins (NLRs). All these
PRRs have been shown to recognize influenza virus in tissue culture and in vivo and they all seem to work together to clear the influenza infection [59].
These innate immune responses against influenza infection involve cytokine proinflammatory responses and antiviral responses. The proinflammatory cytokine response recruits effector cells that will help to clear infection, and cells needed for adaptive immunity.
When this response is not controlled properly, it leads to severe lung viral pneumonia, seriously complicating infection. The antiviral cytokine response facilitates intracellular control of viral replication using interferons and interferon-downstream signaling. In this response, the quick production of type I interferon is a key component of the antiviral response [60]. Recently, the interferon-inducible transmembrane protein 3 (IFITM3) and S-palmitoylation are reported to mediate the innate immune role of interferons in combating pathogenic influenza A strains [61].
17 | P a g e
As mentioned earlier, the first line of protection against influenza infection are PRRs (e.g.
TLRs, and RLRs), these are activated after recognition of viral moieties (e.g. peptides, glycopeptides and nucleic acids). TLRs are localized on the cell membrane and endosome.
Influenza virus double-stranded RNA (dsRNA) is detected by TLR3 in the endosome and influenza single-stranded RNA (ssRNA) is recognized by TLR7 in human plasmacytoid dendritic cells [62]. After binding of PAMPs to TLRs, the downstream signaling pathway leads to expression of protein effectors and proinflammatory cytokines by innate immune cells. Lack of
TLR is proven to enhance lethality by a failed inflammatory response during influenza infection in vivo [63].
In the cytoplasm, virus is detected by RIG-I and NOD2 [59, 64]. RIG-I binds and gets activated by viral ssRNA, which is uncapped and has a 5-triphosphate moiety. Once active,
RIG-I binds to its adaptor (mitochondrial antiviral signaling protein, MAVS; also known as
CARDIF), and activates downstream signaling that provokes NF-kB activation. Activated NF-kB leads to inflammatory cytokine production and activation of interferon antiviral response [64].
In humans, large complexes called inflammasomes (e.g. NLRP1, NLRP3, and NLRC4) are assembled to activate caspase-1. Active caspase-1 activates cytokine precursors through a cleavage event. During influenza virus infection, proinflammatory cytokine substrates IL-1β and
IL-18 are secreted from infected cells and recruit monocytes and neutrophils into the lung. Using a receptor knockout mouse model, it has been demonstrated viral infection is lethal in mice that the lack of IL-1 and IL-8, suggesting that these two cytokines are important for control of viral infection and tissue pathogenesis. Previous studies suggest that NLRP3 inflammasome plays a critical role in murine innate responses to influenza virus, leading to impaired lung inflammation that correlates with increased mortality and defects in virus clearance [65, 66]. Additionally,
Apoptosis-associated Speck-like protein containing a CARD (ASC) and caspase-1, NLRP3 inflammasome components, were shown to be essential for mouse survival and in vivo host
18 | P a g e
immune responses to influenza. However, some reports show that NLRP3 was not required for protective immunity against influenza infection [67].
2.6 Epidemic and pandemic influenza
Unlike other viruses, influenza virus affects all age groups and can recur in any pre-exposed individual. Influenza is usually associated with high mortality in the young and the elderly.
Immuno-deficient patients and pregnant women also show more severe disease and significant mortality. During an influenza epidemic, influenza activity reports increase in a particular location
(e.g. city, town, or country) and it is usually associated with seasonal influenza strains. In temperate climates, epidemics of influenza occur every year but with substantially different rates and severity of disease. In subtropical regions, influenza epidemics do not show established yearly patterns. During an influenza pandemic, the influenza activity is not restricted to a location and is associated with the appearance of new strains usually to which the population has no immunity or pre-exposure. The 20th century experienced three pandemics (1918, Spanish flu
A/H1N1; 1957, Asian flu A/H2N2; 1968, Hong Kong flu A/H3N2), all of which were associated with a high mortality rate. The recent pandemic in March 2009 started in Mexico as an outbreak of a new strain of H1N1 influenza A and rapidly spread to a pandemic classification, as designated by the World Health Organization (WHO) in June 2009; although, with lower mortality rates than previous pandemic strains [68].
2.6.1 Antigenic Drift
Antigenic drift refers to minor antigenic changes usually on influenza HA or NA during its replication and caused by infidelity in RNA polymerase activity. Antigenic drift most commonly affects HA resulting in a gradual accumulation of amino acid changes on the major antigenic sites of this protein. Therefore, antibodies generated after exposure to the previous strains
(before antigenic drift) will not be able to neutralize the antigenic variants (after antigenic drift) as effectively. Gradually, the new variant will be the predominant virus in an epidemic; since this new variant will have a better change to escape from the pre exposed immune response. New
19 | P a g e
antigenic variants can be produced in vitro by exposing the virus-infected-tissue cultures to limiting amounts of antibody [7]. Although antigenic drift of NA is not studied intensively, some reports showed that there are amino acid substitutions in the antibody epitopes in the NA.
Antigenic drift of nucleoprotein leads to an effective escape from cytotoxic T lymphocytes recognition [69].
From studies on HA gene sequence, different types of influenza viruses (type A, B, and C) reveal that isolates from successive years show different pattern of evolution suggesting different antigenic drift patterns. For influenza A, a single lineage circulates in humans resulting in a linear accumulation of mutations in the HA [4, 69]. Therefore, each new strain replaces the previously circulating strain. On the other hand, influenza C has multiple lineages that co- circulate and does not show a linear accumulation of mutations on HEF gene [4]. Evolution of influenza B viruses lies in between the evolution of influenza A and C, since two major lineages co-circulate worldwide (e.g. Yamagata and Victoria lineage). Usually, for all influenza HA, single point mutations do not significantly contribute to increase the antigenic distance; but when two or more mutations on the HA binding sites occur, they increase the antigenic drift considerably.
2.6.2 Antigenic Shift
Antigenic shifts in pandemic influenza usually results from the combination of two or more different strains of influenza to form a new subtype having a mixture of the surface antigens of original strains (e.g. H3N2 + H5N1 → H5N2). This virus is virtually new to the population, which has not been exposed previously [7]. Almost no serologic correlation exists between the new HA or NA antigens (after antigenic shift) and the old viruses (before the antigenic shift). By consensus, these viruses receive different designation in nomenclature [4, 5].
Figure 1.7 shows how antigenic shift and antigenic drift relates to population immunity. If a virus (HxNx) is introduced into a population, antibodies against this new virus will be lacking, resulting on an influenza pandemic. After the pandemic, the population will develop immunity.
This favors the emergence of viruses with antigenic changes in the HA or NA (antigenic shift),
20 | P a g e
which will spread throughout the partially immune population. This phenomenon is repeated with subsequent epidemics and it is represented in Figure 1.7 as many small red peaks. Throughout the years, the immunity of the population to all circulating variants increases. This favors the conditions for the spread of a new pandemic virus (HyNy, antigenic shift) to which antibody is lacking in the population [4].
Figure 1.7. Schema of occurrence of influenza pandemics and epidemics in relation to level of immunity in the population. HxNx and HyNy represent influenza viruses with completely different HA and NA. (Modified from Treanor J (2009) [4]).
3 Developing diagnostic test
The objective of Chapter 2 and 3 of this study is to develop a rapid diagnostic test, which is able to detect influenza. In this section, we discuss the advantages and disadvantages of current diagnostic tests considering their turnaround time, cost, complexity, and theoretical basis. In addition, we also discuss the need of a new approach for influenza detection.
Both influenza antigenic drift and shift are the cause of annual epidemics and sporadic pandemics, respectively. Influenza infection complications can be life threatening especially to very young and elderly people. Vaccination is the main defense mechanism for preventing annual influenza epidemics; while, antiviral drugs are recommended for prophylaxis and treatment. Diagnosis, isolation and identification of circulating influenza strains are needed for
21 | P a g e
the annual preparation of flu vaccines. Rapid influenza diagnosis tests (RIDT) are shown to reduce antiviral use. However, current influenza diagnostic tests encounter influenza antigenic drift and shift and need to be re-assessed annually [70]. Moreover, the variability of RIDT sensitivity and specificity during different epidemiologic influenza cycles highlights the need to improve these assays [71, 72]. Current influenza diagnostic tests include: viral culture, rapid antigen testing, serology, immunofluorescence assays, and polymerase chain reaction (PCR). It has been shown that the sensitivity and specificity of any influenza test is also affected by laboratory performance and the type of test or patient sample [68].
3.1 Virus cell culture
Virus cell culture is the gold standard for influenza diagnosis. However, this technically demanding work requires the maintenance of several cell lines and the turnaround time is 10 days. Consequently, this expensive diagnosis needs laboratory expertise, additional reagents and equipment that limits its use to a laboratory setting [68]. Influenza can also be diagnosed by rapid cell culture assays, which consist of the evaluation of viral antigens in the infected cell culture by immunological methods. Those results can be obtained in 18-40 hours. However, rapid cell culture assays are still restricted to laboratory use [73].
3.2 Serology
Virus culture is usually more rapid than serological diagnosis. These methods require a significant increase in antibody titer to demonstrate acute infection [68]. However, serological diagnosis is used when clinical samples are not adequate, when laboratories cannot perform virus isolation or when virus detection is negative. In addition, these methods are used in epidemiologic studies to determine exposure to different influenza strains.
3.2.1 Hemagglutination inhibition (HAI)
HAI test helps in identifying the type of influenza viral isolates or evaluation of antibody response to vaccination. HAI test is widely use in epidemiological and immunological studies
[74]. This method examines influenza HA protein ability to bind to and agglutinate red blood cells
22 | P a g e
(RBCs). If anti-HA antibodies block HA, RBC agglutination does not occur. This is done with yearly prepared reference antisera that contain antibodies against circulating influenza viruses or with patient’s antibodies that are supposed to neutralize HA binding. Reference antisera are derived from wild type strains or a strain with equivalent antigens [4]. Consequently, both antisera (reference and patient’s) are affected by antigenic drift and shift. These methods are not rapid enough results to influence clinical decision making. Some other disadvantages of the
HAI test include: the removal of nonspecific hemagglutination inhibitors in sera, the standardization of references, and the need for laboratory expertise. In some cases, the influenza infected patient’s sera might not have neutralizing antibodies giving a false negative result. Despite its disadvantages, this test is still used for global influenza surveillance, characterizing the antigens of influenza isolates, and it is preferred over other diagnostic tests
(e.g. complement fixation and ELISA). WHO influenza Reagent Kit Reagents, which is a HAI test, is provided for the diagnosis of influenza H1, H3 or type B HA infections [68]. Other methods neutralize influenza viruses with serum antibodies during infection of cell cultures, which in combination with an ELISA can give results in 2 days. Neutralization assays are highly sensitive and specific for detection of human neutralizing antibodies. However, these methods are arduous with a long turnaround time.
3.3 Reverse-transcription polymerase chain reaction (RT-PCR)
The RT-PCR amplified and detected viral RNA by reverse transcription. Molecular identification of influenza can be used directly on patient’s samples giving rapid results during outbreaks. Results can be available in hours and can be also used for the identification of an unknown influenza strain and monitoring genetic evolution of influenza viruses. However, these tests need to be performed on a BSL-2 laboratory with careful manipulation of samples and following certain regulations [68]. Moreover, RT-PCR primers are designed based on the sequence of highly variable HA, NA or M1 genes. Since all those genes go under antigenic drift
(especially HA and NA), newly formed point mutations on the gene sequence can prevent the
23 | P a g e
binding of these primers, leading to a wrong diagnosis. Moreover, like virus cell culture, RT-PCR is also an expensive technique that needs laboratory expertise skills.
3.4 Neuraminidase activity
NA activity can be measure to detect influenza. Current assay use highly sialylated fetuin protein or small substrate molecules that mimic NeuAc. Antisera might also be used to block NA activity in these assays. Since the emergence of NA-inhibitor resistance viruses, a NA inhibitor susceptibility surveillance is needed. Centers for Disease Control and Prevention (CDC) have developed a chemiluminescent NAI assay to assess the influenza susceptibility to NA inhibitor zanamivir and oseltamivir; while National Institute for Medical Research in the United Kingdom uses fluorescent substrate. This assay is quick; but, requires a careful manipulation of samples, additional reagents, equipment as well as expertise skills [68].
3.5 Immunofluorescence
Immunofluorescence antibody (IFA) staining of clinical samples or tissue culture is fast and sensitive to detects influenza. IFAs are usually point-of-care tests and some can differentiate influenza A and B. This method uses commercially-available monoclonal antibodies against viruses. Due to the variability of influenza antigens, many commercial rapid tests for influenza diagnosis greatly differ in specificity and sensitivity. WHO recommends that these assay results should not be considered alone. One advantage of IFA is that it can provide results within 30 minutes. However, some disadvantages are: the complex procedure and requirement of additional equipment (e.g. fluorescence microscope). In addition, the assay sensitivity is greatly affected by the sample quality, antibody specificity, and expertise skills.
3.6 Rapid Diagnostic Tests for influenza (RIDT)
RIDTs detect influenza using monoclonal antibodies against viral antigens [72]. The most common antibody used is α-nucleoprotein antibody, since nucleoprotein is abundant in both influenza type A and type B virions. Consequently, the sensitivity of these RIDTs rely on cross- reactivity of antibodies with the viral antigen. The turnaround time of this test is very quick
24 | P a g e
(around 15 minutes). More than ten RIDTs are approved to use in USA. Some can distinguish influenza A and B while others cannot. Even though they can use several types of patient’s samples, the test accuracy can vary based on the quality of the samples. The sensitivity (50-
70%) and specificity (90-95%) of each test varies if compared to viral culture or RT-PCR [75]. To have accurate diagnosis, it is recommended that sample collection should be very close to development of initial symptoms. During low influenza activity (Spring-Summer), false-positive and true-negative test results are more likely to occur; while during high influenza activity
(Winter), false-negative and true-positive test results are more likely to occur [76].
A recent retrospective study demonstrated that patients with confirmed pandemic
A/H1N1/2009 infection and symptoms did not have a positive RIDT result [77]. Moreover, studies with swine-origin H1N1 viruses demonstrated a variable sensitivity within 20 RIDTs in
Japan. Some tests lack sufficient sensitivity to detect low-virus load during the initial infection
[78]. Pandemic H1N1/2009 virus detection by USA-approved RIDTs was also found to have low sensitivity in comparison to RT-PCR [79].
3.7 Pyrosequencing
Pyrosequencing is a relatively new method to detect influenza. It is based on the sequencing of genetic material by synthesis [80]. It uses the pyro-phosphate (PPi) produced during synthesis of the PCR product of interest to detect the nucleotide incorporated at each time. This method performs real-time nucleotide sequencing for typing and detection of pathogens and also single nucleotide polymorphisms in pathogen genes. Pyrosequencing is reported to successfully screen adamantine-resistance influenza viruses by analyzing M2 genes [81].
Moreover, this technique is also used to detect single point mutations associated with NAI- resistance [82]. Recently, this technique was also used for subtyping a large number of influenza A viruses [83]. The advantages of the pyrosequencing assay includes: high sensitivity, high specificity, ability to determine antiviral susceptibility, and the ability to be use in high throughput screenings. However, to detect NAI-resistant influenza strains, pyrosequencing
25 | P a g e
assays use already published genetic markers of antiviral resistance (e.g His274Tyr). Therefore, new antiviral-resistance point mutations, which are described every year, will be overlooked.
Consequently, pyrosequencing will not replace NA inhibition assays and NAI susceptibility assays due to novel molecular mechanisms that confer NAI-resistance. In addition, this technique is highly expensive and requires laboratory expertise skills to perform the assay and data analysis.
3.8 Need for a diagnostic test
Current diagnostic tests (e.g. RIDTs) show a deficient sensitivity for influenza detection.
Other tests (e.g. RT-PCR, HAI, and virus cell culture) have a high cost, need laboratory expertise or the turnaround time cannot influence clinical decisions. During the recent pandemic
H1N1/2009, influenza diagnosis was inaccurate [77-79]. Moreover, some patients did not receive any treatment because of negative results (false negative). The delays in treatment increase the infection severity and mortality. Taken all together this highlights the importance of accurately detection of influenza and to determine antiviral-susceptibility of infecting strains, in order to improve influenza treatment and disease outcome. There is also a need to develop more sensitive RIDT to detect both influenza A and B in respiratory samples at all stage of disease allowing the opportune implemention of outbreak-control interventions. Moreover, an accurate RIDT with the ability to distinguish NAI-resistant strains will improve the NAI use by treating patients only infected with NAI susceptible strains. This will eliminate unnecessary treatments and the inadequate use of antivirals.
In this study, we used a new approach to develop an influenza test. We used two different invariable characteristics of the influenza virus: glycosylation patterns on surface glycoprotein
(HA and NA) and neuraminidase activity. Both characteristics are not subject of antigenic drift or shift; therefore, new seasonal or pandemic strains will be detected. We captured virus by using molecules that specifically recognize the viral glycosylation pattern on glycoproteins, and we used NA enzymatic activity for virus detection. In the next section of Chapter 1 (section 4), we
26 | P a g e
discuss the features of HA and NA glycoproteins, as well as, NA inhibitor resistance. In Chapter
1 section 5, we focus in the capturing molecules needed to detect influenza. Our results of the characterization of influenza NA activity is detailed in Chapter 2. Our results with capturing assay are detailed in Chapter 3.
4 Major influenza antigens
4.1 Hemagglutinin binding preferences
As mention before in Chapter 1 section 2.3, Hemagglutinin (HA) is a trimer formed by two domains: a globular and a stem domain (Figure 1.3) [10]. All three monomers of functional
HA together form a principal α-helical coil and the three globular heads, which each contains a
NeuAc-binding site (Figure 1.3) [84, 85]. Each of the globular domains is formed only by HA1 with highly variable loops and eight β-sheets. These domains not only form the receptor binding sites but also the antigenic epitopes.
4.1.1 Receptor binding site and binding preferences
It is well established that NeuAc found in glycoproteins and glycolipids is the host receptors for influenza virus [10]. This terminal NeuAc binds to conserved residues at the top of the HA molecule on each monomer; this region is conserved throughout antigenic variation
(Figure 1.3). Many structural studies agree that the binding of NeuAc, receptor analogs or other sialylated sugar chains orientate very similarly when bound. One side of the pyranose ring of the
NeuAc faces the base of the binding site [84]. Other side groups of the sugar such as carboxylate, acetamido nitrogen, and hydroxyl groups form hydrogen bonds with conserved side-chain in the molecule. These interactions of HA binding site and substrate have also been probed by site-directed mutagenesis of residues in the receptor-binding site [10].
Once the trimer is assembled, a second NeuAc-binding site is found in between the monomers of the HA trimer. It is located in the interface of contact of two HA1 domains and one
HA2 domain; these second binding site will be exposed when the two domains dissociate in the membrane fusion conformation [10]. This binding site has fourfold weaker affinity for α2,3-
27 | P a g e
sialyllactose than the primary site, and the physiological importance of the second binding site is still unclear [86, 87].
There are two major linkages between NeuAc and the galactose residue of glycan chains naturally found in glycosylated proteins: NeuAc(α2,3)Gal and NeuAc(α2,6)Gal. Different HAs show different binding preferences for each linkage. For many years, this linkage was believed to determine whether an influenza virus binds to avian (α2,3) or human cells (α2,6) [88-91]. A point mutation in HA1 is sufficient to switch binding preferences from α2,3 to α2,6 or vice versa
[92, 93]. This single point mutation is reported to be on HA binding site (Leu226Gln on A/Hong
Kong/68 HA) and it can change the preferences from α2,6 to α2,3. However, the residue 226 does not directly contact the substrate but it changes the conformation of the receptor-binding pocket [10]. Moreover, South Carolina HA (A/SC/1918) is reported to have a substitution in residue 225 from Gly to Asp, which is sufficient to switch receptor preference from α2,3 to α2,6 in a cell-based assay [94].
Conversely, influenza virus A/NWS/33 is also reported to be able to bind to α2,8 linked
NeuAc present on ganglioside glycolipid [95, 96]. Additionally, H2 viruses preferentially bind
NeuAcα2,3Lactose and the human H5N1 isolates seem to prefer to bind α2,3 receptors [97, 98].
Glycan array studies reveal that two 1918 H1N1 variants, A/SouthCarolina/1918 and
A/SouthCarolina/1918 (K222L) both with Asp190 and Asp225 bound exclusively α2,6 receptors
[94]. On the other hand, the 1918 HA of A/NewYork/1918 variant also with Asp190 and Asp225, and Gly at position 225 had equal preference for both α2,6 and α2,3 receptors [94]. When a mutation Asp190Glu in the New York variant is present, the HA receptor preference is restored to the avian strains (α2,3) [94]. This suggests that a single point mutation can change the receptor binding specificity.
The dogma states that human HA will only bind α2,6 glycans and the avian HA will only bind α2,3 glycans; however, recent data suggest this is not true [99, 100]. In vitro studies with human epithelial cells from air tract show that human cells express NeuAc α2,3 sugars as well
28 | P a g e
as NeuAc α2,6Gal [101]. Moreover, glycan array studies performed in the recent pandemic
H1N1 isolates (A/California/4/2009 and A/Hamburg/5/2009) and the seasonal H1N1 virus
(A/Memphis/14/96-M) reveals that pandemic viruses bound to the majority of α2,6 glycan receptor irrespectively of the backbone chain length (up to 6 residues) or type (linear or branched) and they are also able to bind α2,3 receptors [102, 103]. In contrast, the H1N1 seasonal strain bind exclusively to α2,6 long saccharide sequences and not to any of α2,3 glycans [104]. These data suggests that avian seasonal H1N1 has now changed preferences in binding human cells α2,6 sialylated receptors.
A recent study on the receptor binding specificity of HA of H3N2 viruses shows diversity in binding glycan receptors over 20 years of isolates [105]. The pattern of binding saccharide sequences remarkably change year to year. 1997 isolates only prefer to bind longer sugar chains; while two 1996 isolates (A/Wuhan/359/95-like and A/Sydney/5/97-like) bind only α2,3 sulfated glycans. The 2003-2006 isolates also bound α2,3 sequences but with lower affinity than
α2,6 sequences. Surprisingly, isolates of year 2010 prefer to bind to α2,3 over α2,6 glycans and their changes in binding specificity do not affect infectivity at all [105].
There are other factors that affect binding of HA to glycan residues. Previous data show that the affinity of binding of A/NWS/33 virus to NeuAcα2,3Lactose is 10-fold higher than the affinity to bind NeuAcα2,3Lactosamine and 3-fold higher than the affinity to bind
NeuAcα2,6Lactosamine [106]. A point mutation in residue 227 from Pro to His of HA results in a decreased affinity for NeuAcα2,3Lactose, without changing NeuAcα2,6Lactosamine binding affinity. These data show that additional groups in the glycan chain residues will also affect binding of HA. Furthermore, binding of HA is demonstrated to be affected by factors that include structure, topology and density of glycans [107]. The length of the spacer used to attach the glycans, the spacer hydrophobicity, and the glycan chain architecture (mono vs biantennary) affect binding [107].
29 | P a g e
Glycosylation of HA also affects binding to host glycans receptors [108] and helps it escape from detection by host immune system [109]. Different subtypes of influenza A HA show great variation in the glycosylation sites of HA1; whereas glycosylation of HA2 is more conserved. It has been shown that glycans near the cleavage site of HA modulate its activation and consequently, its infectivity. Mutational deletion of HA glycosylation sites can affect binding to host receptor [110]. Wang et al. (2010) constructed four influenza-HA variants with different glycosylation forms (typical N-glycans, NeuAc-free form, high-mannose form, and NeuAc
Glucose form) to evaluate the role of HA glycosylation on binding affinity and specificity using a synthetic NeuAc microarray [111]. Shorter forms of the N-glycan chains in the shorter on HA increased NeuAc binding affinities. However, these forms decreased specificity toward different
NeuAc ligands [111]. In addition, Liao et al. evaluated binding specificity of several HA subtypes and virus to a library of sialosides finding that binding is affected by glycosylation at Asn27 residue, which seems to be essential for receptor binding [104].
Glycosylation of HA affects host immune response [112, 113]. Interestingly, Das et al. found that influenza escape from immune system by glycosylation [114]. His data support that glycosylation on residue 131 decreases viral binding to cellular receptors and additional substitutions in HA are needed to increase receptor avidity. Additionally, Wei et al. found that glycans close to antigenic epitopes inhibit antibodies directed to H1N1 seasonal strain or H1N1 pandemic strains (1918 and 2009) [110].
Recently, the receptor binding specificity of influenza virus HA has been demonstrated to have an affect on the host innate immune response independent of viral replication or infectivity
[113, 115]. Two recombinant mutant influenza viruses from highly pathogenic H5N1 were generated with different receptor specificity, α2,3 or α2,6 NeuAc, and the immune response was evaluated in vitro. Viruses with preferential α2,3 NeuAc affinity increased levels of proinflammatory cytokines and interferon-inducible genes in human dendritic cells as compared with viruses with α2,6 NeuAc binding specificity. This study suggests that the strong host
30 | P a g e
inflammatory responses triggered by H5N1 may be due to its α2,3 receptor specificity [116].
Moreover, it is possible that two non-mutually exclusive mechanisms for sensing the viruses might exist on human dentritic cells and macrophages wherein α2,3 NeuAc-binding strains are sensed distinctly from α2,6 NeuAc-binding viruses, the former inducing a rapid and enhanced proinflammatory response, possibly due to distinct signaling cascades of activation in immune cells [115, 116].
In this study, we used glycosylation patterns on HA to capture influenza virus. We also took advantage of the HA binding preferences and we captured virus using a highly sialylated mammalian protein, fetuin (Figure 1.8), which has α2,3 and α2,6 NeuAc terminal linkages in tri- tetra-antennary sugar chains [117]. This protein is previously reported to be able to bind to HA proteins and virus [118]. We combined these intrinsic and invariable characteristics of HA with an antibody-free detection assay, NA activity assay. This rationale of this NA activity assay to detect influenza is discussed further in Chapter 1 section 4.2. Our results obtained with capturing assay to detect influenza are detailed in Chapter 3.
Figure 1.8. Example of Structure poly antennary glycan chains found on fetuin. NeuAc are shown as purple diamond, galactose as yellow circle, NAc glucose as blue squares and mannose residues as green circles. Figure generated using Chemdraw.
4.2 Neuraminidase
Neuraminidase (NA) is one of the major membrane glycoproteins found on the surface of influenza. NA, also called sialidases, specifically catalyze the cleavage of terminal sialic acid
(NeuAc) from viral and host glycoconjugates helping virus leave the infected cell and ensuring that the virus does not get stuck on the cell surface [119, 120]. A fully functional NA forms a
31 | P a g e
homotetramer that cuts off α2,3, and α2,6 glycosidic linkages of terminal NeuAc residues in oligosaccharides, glycoproteins, glycolipids, and synthetic substrates. NA is type II membrane protein with a single-pass through the membrane and a transmembrane domain involved in lipid raft association during viral intracellular transport [121].
Ten serologically distinct subtypes of influenza NA have been described. Nine of them belong to influenza A (N1-N9) and one to type B influenza viruses (type B NA). Interactions with antibodies define NA subtypes; thus, subtypes that are all neutralized by a similar set of antibodies will belong to the same subtype [4]. NA subtypes seem to be species-restricted and plays a major role in effectiveness of influenza infection. Only subtypes N1 and N2 have been positively linked to epidemics in human. While both are from influenza A viruses, they belong to two phylogenetically and structurally distinct groups. Group-1 contains N1, N4, N5 and N8; while group-2 contains N2, N3, N6, N7 and N9 [85]. The structure of group-2 NA of influenza A virus was used to model and design current drugs (e.g. oseltamivir and zanamivir) using the conserved structure of the NA active site.
4.2.1 Receptor binding site
Analysis of the NA active site shows conserved residues in all NA subtypes. These conserved residues are: residues in catalytic sites (Arg118, Asp151, Arg152, Arg224, Glu276,
Arg292, Arg371, and Tyr406) that directly interact with substrates, and residues that support the catalytic residues (Glu119, Arg156, Trp178, Ser179, Asp/Asn198, Ile222, Glu227, His274,
Glu277, Asn294, and Glu425) [40, 122-124].
In the active site of influenza A NA, three arginine residues, (Arg118, Arg292 and Arg371) directly bind to the substrate carboxylate group (Figure 1.9). Glu276 and Glu277 forms hydrogen bonds with two hydroxyl groups at carbon 8 and 9 of the NeuAc. Another Arg 152 interacts directly with the acetamido group of NeuAc [40, 122]. Calcium ion coordinates amino acids near the substrate binding and it is believed to maintain substrate binding conformation
(Figure 1.9).
32 | P a g e
Other roles of NA have also been described during initial stages of infection [125, 126].
Recently, Matrosovich et al. showed that NA has an essential role for the initiation of influenza virus infection in ciliated epithelium of human airways [126]. The NA role at this initial stage is to remove NeuAc from mucins, cilia, and cellular glycocalix. Binding of influenza to these decoy receptors prevents it from binding to receptors capable of promoting host cell entry. It has been shown that treatment with NA inhibitors at initial steps of infection can reduce the efficiency of the viral infection, without significantly affecting virus binding [125]. Therefore, it was suggested that NA inhibitors interrupt the normal virus entry step.
Figure 1.9. Structure of the influenza N2 (PDB 2BAT) bound to Sialic Acid and Ca2+. Substrate binding residues are highlighted in green. NeuAc is the substrate and is shown in pink. High affinity Ca2+ binding site is located near the substrate binding site. Calcium ion is shown as a yellow sphere and its five coordinate oxygen groups are shown in red. Figure generated using PyMol.
4.2.2 Calcium effect on influenza NA
It has been known for years that calcium played a role in influenza neuraminidase activity.
Research suggests that the presence of the calcium ion contributes to thermostability, allows the enzyme to continue functioning at high temperatures, and increases its activity [127]. In this study, we explored the effect of calcium on NA activity of different NAI-sensitive and NAI- resistant influenza strains along with parainfluenza and Streptococcus strains. Our results of the characterization of influenza NA activity is detailed in Chapter 2.
33 | P a g e
Carroll et al. provided evidence that Ca2+ ions are able to stimulate N1 activity (A/WSN/33,
A/PR/8/34, and A1/FM/1/47) up to 100 fold more [128]. Other metal ion salts (Mn2+, Mg2+, and
Fe2+) can also increase N1 activity, but Ca2+ has the maximal effect. Other cations, Zn2+ and
Hg2+ ions or chelators can inactivate N1. Similar results were obtained by Kiyotani et al. (1987) with avian NAs (N1, N2, N3, N6, and N7) but not with type B NA [129]. The latter seems to not have a Ca2+ dependent activity. Additionally, Chong et al. demonstrate that Ca2+ exerts a specific effect on Vmax/Km leading to an increased rate of interaction between NA and NeuAc
[130]. In 1991, Varghese et al. reports that the crystal structure of the influenza A virus NA, had putative Calcium binding sites [131]. Later, in 1994, Pascal et al. and other reports that the structure of the influenza B virus NA, head has two distinct Ca2+ binding sites [123, 124]; similar
Ca2+ binding sites are found for Influenza A NA [132-134]. One is located near the NeuAc binding site and has high affinity for the metal while the second one has low affinity for Ca2+, which is located in the molecular fourfold symmetry. The Ca2+ at the low-affinity binding site can easily be displaced by other metal ions or removed completely by EDTA. The high affinity Ca2+ site and the NA active site are structurally connected; this is shown in Figure 1.9. Studies on the thermo-stability of influenza B NA at 37⁰C, in the presence and absence of Ca2+, concluded that
Ca2+ stabilized the NA tetramer. Johansson et al., in 2003, demonstrated that the complete removal of all divalent cations by dialysis not only limits enzymatic activity but also significantly reduces NA immunogenicity evaluated by a NA inhibition antibody assay [135].
Recently, Li et at. reported an extra Ca2+ binding site on 2009-pandemic H1N1 NA. This calcium ion is binding residues Asn381, Asn389, Asp387, Ser319, and Asp379. The role of this
Ca2+ is still unclear. The role of calcium in influenza NA activity is studied further in Chapter 2.
Interestingly, Smith et al. reported the structure of a NA (A/NWS/whale/Maine/1/84,
H1N9), which did not contain Ca2+ in the high affinity site in three of the monomers of the NA tetramer [136]. The absence of calcium affects the conformation of residues that directly interact and participate in substrate binding, and provides a basis for understanding of the role of
34 | P a g e
calcium in substrate and inhibitor binding. Surprisingly, in one of the monomers of HA structure reported by Smith et al., two residues, which directly participate in binding the substrate, adopt unusual conformations in the absence of Ca2+. Arg292 folds back toward Asn294 forming a hydrogen bond with the side-chain of Asn294 [136]. In this calcium-free conformation, Arg292 forms hydrogen bonds with a total of three residues: Glu276, Glu277 and Asn294. In the other three monomers, the conformation of the side chain of Arg292 is fully extended, forming hydrogen bonds with Glu277 and Asn294 [136]. Significant oseltamivir resistance is reported for the R292K mutant of N9. Moreover, specific activity of mutant R292K is only 20% of that of the wild type in the absence of drug. It is possible that this mutant requires more free energy to accommodate inhibitors and ligand groups, which need to interact with residue 292 [136]. In this mutant the lysine side chain is not extended to the carboxylate group of these inhibitors or the substrate and lacks two hydrogen bonds normally formed between substrate and residue
Arg292. Previous data show that in the Ca2+-free structure, Arg292 and Glu276 adopt new conformations. If their conformation is maintained upon ligand binding, the number of hydrogen bonds to the ligand will be reduced significantly. This will decrease the affinity of the NA for the ligand. Simulations of Ca2+-bound and Ca2+-free N1 complexes with oseltamivir using molecular dynamics give structural insight into the role of Ca2+ in substrate binding [137]. Using two different force fields, it is shown that Tyr347 of N1 residue works as a clamp to favor binding pose of the ligand. Moreover, in the absence of Ca2+, the free energy required for oseltamivir binding is increased by 5 kcal/mol [127, 137]. In this study, we explored further the role of calcium on influenza NA activity; our results are detailed in Chapter 2.
4.2.3 Neuraminidase inhibitors
For influenza prevention and control, annual vaccination is the recommended method. For the management of influenza outbreaks and pandemics, antiviral drugs are recommended to use as prophylactic and therapeutic medication. Currently, adamantanes and neuraminidase inhibitors (NAIs) are the only two types of anti-influenza drugs approved for treatment in humans
35 | P a g e
[138, 139]. Antiviral resistance development to both drug types is increased due to the ability of this pathogen to adapt rapidly and also because each drug only targets a single stage in influenza life cycle [140].
In contrast to adamantanes that only work against influenza type A, NAIs have proven to work effectively against influenza type A and B. Two NAIs, inhaled zanamivir (Relenza™) and oral oseltamivir (Tamiflu™) are licensed for influenza treatment worldwide (Figure 1.10) [141].
Figure 1.10. Structure of NA substrate and inhibitors. The structures of sialic acid (SiAc), zanamivir, and active form of oseltamivir are shown in sky blue, blue, and yellow, respectively. Figure generated using Pymol.
The influenza NA glycoprotein cleaves the terminal α2,3 and α2,6 NeuAc on cellular moieties. During influenza infection, NA is needed for preventing self-aggregation, provoking release of virions and spreading infection (Figure 1.11). The NA catalytic site located in the globular head of the protein is highly conserved among influenza types (A and B). The NA inhibitors resemble the NA natural substrate, NeuAc. NA inhibitory drugs were designed based on crystal structures of group-2 NAs (N2 and N9), and type B NA and not on group-1 NA. Since group-2 NA and distantly related influenza type B NA have very similar structures at the active sites, it was assumed that NA of group-1 would be also similar [85]. However, recent studies show that group-1 and group-2 NA display different micro domains very close to the active site, while the active site of group-1 and group-2 NA remains nearly identical.
36 | P a g e
A. Neuraminidase activity
B. Neuraminidase Inhibitor
Figure 1.11. Mechanism of Action of Neuraminidase Inhibitors. (A) Action of neuraminidase in the continued replication of virions during influenza infection. (B) Viral replication is blocked by neuraminidase inhibitors preventing virions from being released from the host infected cells. Figure generated using Chemdraw.
After analysis of the structures of N1, N4 and N8 of group-1 NAs, recent data reveals one loop and one 150-cavity contiguous to the active site are found on group-1 and not group-2
NA [142, 143]. That cavity makes it possible to add a 4-amino group of oseltamivir that is suggested to increase the binding NAIs. Therefore, the cavity found near the active site of group-1 NAs could be exploited to develop new antiviral drugs. Recent studies indicate that laninamivir (recently approved for flu treatment in Japan) and zanamivir are more effective against group-1 NA with a 150-cavity than group-2 NA with no 150-cavity [143].
Currently, zanamivir and oseltamivir are approved for influenza treatment and prophylaxis worldwide since the late 1990s (Figure 1.10) [144]. Zanamivir is 4-deoxy-4- guanidino analogue of NeuAc. Oseltamivir is an ethyl ester prodrug, which is turn into a carboxylate form in the liver by esterases enzymes. Zanamivir mimics the NeuAc natural substrate, smoothly fitting into the active site. In contrast oseltamivir has modifications needed to increase its oral availability such as a lipophilic side chain. Compared to the adamantanes, the
37 | P a g e
two licensed NA inhibitors zanamivir and oseltamivir are associated with very little toxicity.
Moreover, NAIs are effective against all NA subtypes of influenza type A and B in vitro and in animal models.
Before the introduction of NAIs for flu treatment, little differences in inhibitory concentration values were found among NAs for both influenza A and B viruses. Studies on
IC50 of NAs demonstrated that influenza B viruses were 3-times more sensitive to zanamivir
(IC50 = 4.2 nM) than to oseltamivir (IC50= 13 nM). A/H1N1 viruses also showed to be slightly more sensitivity to zanamivir (IC50= 0.92 nM) than oseltamivir (IC50= 1.34 nM), while A/H3N2 viruses showed the opposite (oseltamivir, IC50= 0.67 nM > zanamivir, IC50 = 2.28 nM) [145].
Well-designed NAIs are highly effective antivirals; however, evolution under selective pressure unsurprisingly produces resistant strains. After oseltamivir treatment, resistant influenza A strains have been isolated. Resistant isolates derived from H1N1 and H5N1 infections usually have the His274Tyr mutation on NA, while resistant isolates from H3N2 viruses contain Arg292Lys or Glu119Val mutations [15, 42, 139].
The mechanism of resistance to oseltamivir is based on the rearrangement of residues needed accommodate the inhibitor in the active site. Hydrophobic groups of oseltamivir (Figure
1.12) interact with Glu276 residue and require its reorientation to fit into the active site [15, 40,
42, 146]. Figure 1.12 shows how residue E276 needs to rotate towards R224 to form a larger cavity to fit hydrophobic groups of oseltamivir at the active site of NA. Mutations that avoid this rearrangement of Glu276 lead to resistance to oseltamivir [147]. Isolates from patients showed mutations that include His275Tyr, Arg292Lys, and Asn294Ser; all inhibiting Glu276 re-rotation
[148]. The mutation Glu119Val enables the accommodation of a water molecule at the active site which would interfere with oseltamivir binding but will have no effect on zanamivir or sialic acid binding [121].
38 | P a g e
Figure 1.12. Mechanism of resistance to oseltamivir. Rearrangement of E276 in the NA active site is necessary to fit hydrophobic side chain of oseltamivir. Three mutations (R292K, N294S, and H274Y) inhibit this rotation conferring resistance to this NAI. Figure generated using Pymol.
Current circulating seasonal A/Brisbane/59/2007-like H1N1 is reported to be oseltamivir- resistant and it has at least 1 mutation on NA His275Tyr (N1 numbering; His274Tyr in N2 numbering) that confers that resistance [149]. Avian H5N1, pH1N1 and B viruses are still susceptible to the NAIs. However, recently, Hatakeyama et al. reported reduced sensitivity to
NAI of B strains [150]. Mutational studies demonstrate that group-2 NAs His274Tyr had little effect on conferring resistance; while the opposite occurs with group-1 NAs. Moreover, residue
Glu119 is oriented differently in unbound group-1 and group-2 NA, however, that residues interact similarly with oseltamivir inhibitor [134].
Zanamivir resembles more closely the natural NA substrate, Sialic Acid (Figure 1.10) and does not requires a reorientation of residues fitting directly into the active site. Therefore, oseltamivir- resistance mutations do not generate resistance to zanamivir and mutated viruses remain sensitive to this NAI.
Prolonged growth of influenza B in vitro under zanamivir results in Glu116Gly mutation
(119 in N2 numbering). Reduced susceptibility to zanamivir is reported to naturally occur by substitution Gln136Lys in H3N2 and H1N1 viruses. Mutant influenza B strains with Asp197Asn and Ile221Tre substitutions were recovered from oseltamivir- treated patients [151]; while
Arg150Lys mutation was isolated zanamivir- treated children [152]. Interestingly, amino acid
Asp198 does not interact directly with the ligands but forms a salt bridge with Arg152, which directly binds N-acetyl group of sialic acid and oseltamivir. Genetic studies on H1N1
39 | P a g e
demonstrate that Glu119Val mutation provides resistance to oseltamivir, and zanamivir.
Similarly, Glu119Val mutant pandemic H1N1 recombinants show the same resistance. Table
1.3 shows the main mutation that confer resistant to NAIs in influenza A and B after treatment of patients with influenza infections.
Table 1.3. Influenza A and B Mutations conferring resistance to NAIs from clinical isolates Influenza NAI used for Phenotype in NA Inhibition assays† NA mutation* subtype selection Oseltamivir Zanamivir H1N1 H274Y Oseltamivir R S Q136K None S R pH1N1 H274Y Oseltamivir R S H274Y/I222V Oseltamivir R S H274Y/I222R Oseltamivir R I H5N1 N294S Oseltamivir R S H274Y Oseltamivir R S H3N2 N294S Oseltamivir R ? R292K Oseltamivir R ? Δ 245–248 Oseltamivir R I D151A/D None S R Q136K None S R I222V/E119V Oseltamivir R S E119V Oseltamivir R S E119I Oseltamivir R I B R371K None R R I222T None I I D198N Oseltamivir R R R152K Zanamivir R R * Residues on NA sequence (N2 numbering); † S, susceptible; I, intermediate (up to 10-fold increase in 50% IC50 over wild-type); R, resistant (>10-fold increase in IC50 over WT); ?, not described
The structure-based design of zanamivir, which closely resemble the natural substrate, decreases the probabilities to induce resistant mutation without interfering with enzyme function.
On the other hand, oseltamivir has many modifications from sialic acid, increasing the chances to develop resistance to this drug than to zanamivir. Recent studies show that seasonal oseltamivir-resistant A/Brisbane/59/2007-like H1N1 with His275Tyr mutation, contained additional NA changes (Arg222Gln, Val234Met and Asp344Asn) [149]. Recombinant mutant strain with His275Tyr and also a reverse mutation had significantly lower lung viral titers and reduced in vitro replication kinetics, as well as, NA activity when compared to wilt type
40 | P a g e
(His275/Arg222) and the seasonal H1N1 (Tyr275/Gln222). Both the seasonal H1N1 and the wild type have similar infectivity and transmissibility [149]. This data suggest that Gln222Arg mutation compromised seasonal H1N1 virus in vitro and in vivo. In contrast, the mutation
Arg222Gln on NA present in wild type seasonal H1N1 may have stabilized the emergence of this NAI-resistant NA.
5 Capturing of viral particles
As mentioned earlier, the approach of this study to develop a new influenza diagnostic assay which uses capturing molecules that specifically recognize the glycosylation pattern on viral glycoproteins. In this section, we further describe the rationale of this approach, the capturing molecules as well as the possible false negative.
5.1 High manose oligosaccharides
Oligosaccharide processing in the ER is highly ordered, and each step dependent on the previous one. Figure 1.13 shows the glycosylation pathway from the ER to the Golgi apparatus.
Mannose trimming starts in the ER by membrane mannosidases, which remove of one specific mannose residue leading to a 8- or 7-mannose oligosaccharide core [153]. The remaining mannose trimming steps occur in the Golgi (Figure 1.13, step D). HA will escape from the ER before the trimming of mannose residues at the Golgi apparatus. Recent siRNA studies have described the important role of vesicular transport complex, coatomer 1 (COPI), in the HA surface expression; inhibiting COPI function interferes with secretion and trafficking of HA protein to the cell surface [154]. It is possible the HA is using the mechanism to escape from the
ER, since COPI transport proteins back and forth from the cis-Golgi to the rough ER [153]; however, this need further investigation.
High mannose glycosylation patterns (Figure 1.13, step B) are characteristic of viral proteins. The high mannose core of influenza HA have been describe earlier, where different amino acid residues where found to have these core [155, 156]. Figure 1.13 shows the high mannose core found on HA. Mammalian proteins will lack this high mannose core since the will
41 | P a g e
follow in the normal secretory pathway. High mannose core is recognized as pathogen- associated molecular pattern (PAMP) and play an significant role in innate immune response toward viral infection [157].
We used this HA property, HA high mannose glycosylation pattern, as a new approach to capture virus using capturing molecules with specific carbohydrate recognition domains [158].
Our results obtained with capturing assay to detect influenza is detailed in Chapter 4.
Figure 1.13. Glycosylation pathway. N-linked olicosaccharide starts in the ER. After glucose trimming in the ER (Step A), proteins with high mannose glycan (Step B) are processed by glycosidase and mannosidases (Step C). Later, proteins are transported to the Golgi apparatus by vesicular transport complex (COPII and COPI). Further processing and trimming (Step D) yields to terminal glycosylation (Step E). Figure generated using Chemdraw.
5.2 Dendritic cell-specific ICAM grabbing non-integrin (DC-SIGN)
As mention before, HA glycoprotein is folded in the ER by chaperones where it also gets glycosylated by adding high mannose sugar chains. Then, HA viral proteins escape from the ER before the trimming of the mannose residues. In this study, we used DC-SIGN and its analogs to capture influenza virions and viral glycoproteins. In this section, we discuss further the properties of this molecule.
DC-SIGN, or CD209, is a type II transmembrane protein and belong to Group II C-type lectins. C-type lectin superfamily is a large group of carbohydrate-binding proteins, which require calcium for binding. All C-type lectins have C-type lectin domain (CTLD) responsible for binding
42 | P a g e
sugars. DC-SIGN and DC-SIGN protein family have four parts: an extracellular C-terminal
CTLD, a stalk, a single transmembrane domain and a cytoplasmic tail. The stalk is formed by short sequence repeats and mediates protein polymerization to a tetramer. The cytoplasmic tail is believed to transmit signals upon ligand binding (Figure 1.14) [159]. Initially, DC-SIGN expression was believed to be exclusively on dendritic cells; now it is known that DC-SIGN analogs are expressed on endothelial cells, activated B cells, skin dermis, placenta, intestinal and genital mucosa, lymphoid tissues and macrophages [160].
Figure 1.14. DC-SIGN structure model. The tetrameric protein is shown. Modified from available resources at CFG Paradigm Pages
DC-SIGN is reported to interact with several pathogens such as HIV-1, Mycobacterium tuberculosis, Candida albicans, Helicobacter pylori, Yersinia pestis, Ebolavirus and cytomegalovirus glycoproteins, Leishmania sp., Klebsiella pneumonae, Neisseria meningitides,
Dengue virus and Aspergillus fumigatus [160-167]. DC-SIGN is able to recognize mannose branched structures and terminal di-mannoses residues [168, 169]. The tetrameric structure of
DC-SIGN increases the specificity for high mannose sugar chains [169].
Recent data demonstrate that DC-SIGN and other mannose binding receptor have also been shown to interact with influenza. By using cell capturing assay, Wang S-F et al. demonstrated that DC-SIGN mediates infections by acting as a capturing molecule for H5N1 virus [170]. Moreover, using cells expressing DC-SIGN or L-SIGN analog, Londrigan et al
43 | P a g e
demonstrate that these receptors bind to influenza and mediate their internalization by a NeuAc- independent mechanism [171].
Other lectins have been shown to interact with influenza virus. Sato el al. used algal lectin
KAA-2 from Kappaphycus alvarezii, which exclusively binds to high mannose type N-glycans, to demonstrate that this lectin directly binds to viral HA by ELISA and interferes with virus entry
[172]. In addition, Sato et al. also showed that green alga Boodlea coacta lectin, which recognize terminal mannose, directly binds to HA protein from clinical isolate including pandemic
H1N1/2009 [173].
Other mannose binding receptors have also been shown to interact with influenza.
Reading P et al. showed that mannose receptors in macrophages are the major endocytic receptors that mediate entry [174]. Moreover, using a microtiter capture assay, Man T demonstrated that mannose-binding lectin is able to directly bind to H9N2 and H3N2 [175]. More recently, Upham J et al. reported that macrophage receptors –mannose receptor and galactose- type lectin– are needed for influenza A entry into macrophages [176]. However, the role of mannose binding lectin is still unclear; since Eison et al. was unable to demonstrate a clinical association between this lectin deficiency and predisposition to pandemic H1N1 2009 influenza strain in a nested case–control study [177]. Recent studies demonstrate the role of mannose binding lectin in the up-regulation of proinflammatory response during infection of pandemic
H1N1 and H9N2 in vitro [178] and in vivo [179]
Taken together, these data demonstrate that mannose residues in influenza glycoproteins directly interact with carbohydrate binding proteins. In this study, we used that viral property to capture virions and viral glycosylated proteins in order to detect influenza.
In addition to DC-SIGN, we also used DC-SIGN analogs. DC-SIGNR (DC-SIGN related) is a membrane protein that has 77% amino acid homology to DC-SIGN [180]. DC-SIGNR is expressed in liver sinusoidal endothelial cells, lymph node, and placental endothelium; while
DC-SIGN is only expressed on dendritic cells [180]. Mouse DC-SIGN-related protein 1 (DC-
44 | P a g e
SIGNR1), also called CD209b, is membrane protein with one C-type lectin domain. DC-SIGNR1 is reported to be expressed in skin, spleen and lung, dendritic cells and spleen [181].
5.3 Neuraminidases on microorganisms – Possible false positives
Since in this study we captured influenza virus using broad spectrum molecules that recognizes viral mannose glycosylation patterns and we detected influenza using NA activity, it is important to consider other viral proteins or viral particles that could also be captured in this assay. In this subsection, we discuss other viruses and viral glycoproteins that have NA activity and will represent our possible false positive.
Neuraminidases (NAs), which cleave the glycosidic linkages of NeuAc on glycoproteins or glycolipids are widely distributed in among viruses and bacteria and are implicated as virulence factors. Both pathogenic and non-pathogenic bacteria produce NAs. Viruses reported to have
NA activity include members of the family of virus Paramyxoviridae and members of phylogenetic subgroup 2 of coronaviruses.
The majority of the bacteria reported to have NA activity are gram-negative and anaerobic such as Bacteriodes spp. (e.g. B.fragiliza, B. caccae, B. uniforms, B. vulgatus, B. buccaea, B. denticolaa, B. capillosus), Porphyromonas gingivalisa, and Capnophilic ochraceaa [182, 183].
Additionally, gastrointestinal pathogens such as Clostridium sordellii, C. perfringens, S. typhimurium, and B. fragilis also show NA activity. One role of bacterial NA is to remove the
NeuAc residues of secreted salivary glycoproteins weakening the protection they give against invading bacteria. This helps the invading bacteria adhere better, enhancing the survival of bacteria in gut mucosa. Similarly, influenza NA is involved in penetration of the mucus layer in the respiratory tract, helping the virus reach specific host target cell to infect.
Fortunately, several anaerobic bacteria in mouth and intestinal-tract pathogens are unlikely to be found naturally colonizing nasal cavity or in nasal washes of patients with flu-like symptoms. Therefore, NA activity present in nasal washes of patients with flu-like symptoms might work as good indicator of influenza NA presence. One exception to this is Streptococcus
45 | P a g e
pneumonia, which is a NA-producing bacteria and a major cause of bacteremia and pneumonia.
S. pneumoniae is reported to express at least three distinct NAs, among which NanA is highly active and conserved [184]. NanA is expressed in all strains upon interaction with host cells. The pneumococcal NAs are believed to modify host glycoconjugates, expose potential binding receptors and provide a source of sugars. Interestingly, two C-Type lectins (mannose receptor and macrophage-expressed murine homolog of DC-SIGNR, SIGNR1) are reported to bind bacterial capsular polysaccharides derived from S. pneumoniae by distinct binding profiles [185].
Table 1.4. Possible false positive viral human pathogens Membrane NAI Genus Human pathogens Protein susceptibility Rubulavirus PIV 2, 4 HN ? Respirovirus PIV type 1, 3 HN Sensitive Morbillivirus Measles virus H† n/a Pneumovirus Respiratory syncytial virus G* n/a Metaneumovirus Human metapneumovirus G* n/a Severe acute respiratory syndrome Coronavirus HE Resistant (SARS) *, no HA and NA activity; †No NA activity; ?, not described; n/a, not applicable; PIV: parainfluenza virus; HN: Hemagglutinin-neuraminidase; H: Hemagglutinin; G: Attachment Glycoprotein; HE, Hemagglutinin-esterase
Paramyxoviridae members have hemagglutinin-neuraminidase (HN) envelop protein [186], while hemagglutinin-esterases (HE) are found in coronavirues. Both HN and HE have similar roles as HEF of influenza Virus type C, which is able to bind sialylated glycans and cleave the terminal NeuAc. The paramyxovirus HN protein has several roles during infection. It mediates viral entry, viral egression, binding to sialylated receptors, activation of fusion, and cleavage of terminal NeuAc chains [187]. Within the Paramyxoviridae family of viruses, both Mumps virus and Human parainfluenza virus (PIV) have HN protein with HA and NA activity. However, HN has a significantly different structure from influenza HEF and NA. Interestingly, human parainfluenza virus is the third most common cause of respiratory infections; while influenza is the first one. Moreover, human parainfluenza 1 and 3 are mainly responsible for 15% to 20% of all viral respiratory diseases among (infants and children); while human parainfluenza 2 and 4
46 | P a g e
infections are less common and less significant [188]. HN protein of human parainfluenza virus
Type 1 is reported to have two binding sites located in the globular domain of the protein [189,
190]. Both sites bind to terminal NeuAc –α2,3 and α2,8– or sialyl-Lewisx motif [191, 192]. None of the sites recognizes terminal NeuAcα2,6. As recently reported for influenza NA, oseltamivir and relenza inhibit human parainfluenza virus NA activity blocking viral entry. In addition, the use of oseltamivir or zanamivir, during human parainfluenza virus infection, blocks receptor binding and fusion but not viral egression from host cell surface [187]. Interestingly, the inhibitory concentration is in the micro Molar range while for influenza NA are in the nano Molar range
[193]. Reduced sensitivity to influenza NA inhibitor, zanamivir, is reported for Human parainfluenza virus type 3 HN with a mutation at Thr193Ile, which was isolated by serial passages in the presence of NA inhibitor [194]. That mutation increases receptor binding, neuraminidase activity, and reduce sensitivity to zanamivir inhibition. Antiviral drug designed specifically for HN, BCX 2798 and BCX 2855, resulted in a significant reduction in virus particle in the lungs and protection from death, if administered before infection. However, neither drugs prevented death if administered after viral infection [189].
Members of group 2 Coronavirus have NA activity due to the HE protein found on the viral envelop. It is believed that HE gene was acquired relatively recent by lateral gene transfer events form influenza type C [195]. Both HE of Coronavirus and HEF of influenza C have identical acetyl esterase domain; however, the receptor-binding site is changed. Now the ligand binds in an opposite orientation [196]. Group 2 coronaviruses HEs (e.g. human coronavirus;
HCoV and hemagglutinating encephalo-myelitis virus; HEV), bind preferably to 9-O-acetylated
NeuAc and that sugar serves as a receptor. Murine coronaviruses HE esterase use 4-O- acetylated- NeuAc instead while bovine toro virus HEs have a preference for 7,9-di-O- acetylated-NeuAc [197]. Both oseltamivir and zanamivir have no inhibitory effect on cytopathogenicity of Coronavirus in vitro [198].
47 | P a g e
In summary, this study evaluated the binding of influenza virus to develop a diagnostic test.
We used host protein reported to interact with the virus and we also detected virus using NA activity. The capturing proteins used here are natural and synthetic sialylated molecules, carbohydrate binding proteins (Chapter 3). In addition, we characterized and explored condition in which influenza NA activity is optimized while restricting the NA activity of other potential false positive respiratory pathogen (Chapter 2).
48 | P a g e
CHAPTER 2: STUDIES ON NEURAMINIDASE ACTIVITY
1 Abstract
Several human respiratory pathogens, including the viral pathogens type A and B influenza and parainfluenza, and the Gram positive bacterial pathogen, S. pneumoniae, express neuraminidase (NA), an enzyme which cleaves terminal sialic acid moieties from glycoproteins, glycolipids and mucins. We examined the influence of pH, calcium and NA inhibitors (oseltamivir and zanamivir) on NA activity using intact, viable pathogens. The pH optimum for influenza viruses occurs at pH 6-7, while the optimum for parainfluenza and S. pneumoniae occurs at pH
5 and pH 7-9, respectively. Chelation of calcium only caused 50% reduction of NA activity for most influenza isolates; however, an oseltamivir-resistant influenza B strain showed nearly complete loss of NA activity when calcium was chelated. Influenza type A, N1 isolates with the
H274Y mutation, which confers oseltamivir-resistance, displayed levels of drug resistance which varied by more than 17-fold, and low level resistance is not very different from what is seen for stains lacking the H274Y mutation, which are considered to be oseltamivir-susceptible. Overall, influenza B strains were also more resistant to both oseltamivir and zanamivir than influenza A.
Surprisingly, purified recombinant NA proteins displayed different pH optimum, calcium requirements, and high levels of drug-resistance compared to whole viruses expressing the identical NA proteins. These results suggest that inhibitor susceptibility should be performed with intact virus, not recombinant NA. Our results show that measuring NA activity in different conditions can detect and differentiate NA activity of influenza virus, parainfluenza and S. pneumoniae, and may be useful in diagnostic applications.
2 Introduction
Neuraminidase (NA) is an enzyme which cleaves terminal sialic acid moieties, also known as N-acetylneuraminic acid (NeuAc), from glycoproteins, glycolipids or mucins. NA-producing
49 | P a g e
microorganisms are generally found on mucosal surfaces, such as the respiratory tract. NA activity is produced by common human respiratory pathogens [199] which mainly include viruses such as influenza [125] and parainfluenza [200] and bacteria such as Streptococcus pneumoniae [184, 201, 202]. The role of NA during pathogen infection varies. During influenza infection, NA facilitates the spread of the virus, releasing newly synthesized viruses from host cells and from viral aggregates [46, 203]. Recently, influenza NA has also been described to play a role in the initial stage of infection, promoting virus entry by removing decoy receptors on mucins, cilia and cellular glycocalyx [125, 126]. NA is a target for neutralizing antibody, and influenza is constantly undergoing antigenic changes to escape the host immune response.
During antigenic drift, influenza accumulates mutations in all gene segments including surface glycoprotein NA. During antigenic shift, influenza exchanges genome segments from different influenza viruses (e.g. pigs and birds) resulting in a new virus subtype with novel glycoproteins.
Among influenza viruses, ten serologically distinct subtypes of NA have been described. Nine of these belong to influenza type A (N1-N9) and one to type B influenza viruses (type B NA) [4].
The type A influenza NA subtypes can be divided into two phylogenetically and structurally distinct groups. Group-1 contains N1, N4, N5 and N8; while group-2 contains N2, N3, N6, N7 and N9 [85]. Recently, a new structurally and functionally distinct category, N10, has been described [204]. Differences in NA activity appear to play an important role in host selection; while all nine types are functional in avian infection, only viruses expressing subtypes N1 or N2 have established stable lineages to caused human epidemics and pandemics since 1918 [205].
Similarly, since 1980 Yamagata and Victoria have been identified as two distinct genetic lineages of influenza type B. However, these lineages are not categorized as different subtypes of NA [206-208].
Currently, the NA-inhibitors (NAI), oseltamivir (Tamiflu®) and zanamivir (Relenza®), are the only anti-influenza drugs recommended for treatment in humans. NAIs halt influenza infection by limiting replication to only one infectious cycle [138, 189]. In the case of parainfluenza, NAIs
50 | P a g e
acting on hemagglutinin-neuraminidase (HN), the protein with NA activity mediating viral binding and release, not only stop NA activity but also block binding to host receptors and subsequent viral fusion [194]. NAI resistance has emerged in influenza virus, with resistance to oseltamivir more commonly than resistance to zanamivir. While mutations that contribute to NAI resistance confer a selective advantage to the virus in patients treated with anti-influenza agents, NAI- resistant viruses may be less fit in the absence of treatment [209].
Bacterial NAs are also thought to facilitate the utilization of mucins and sialo-glycoproteins as a carbon source. Additionally, desialylated mucins, immunoglobulins, and cell-surface receptors impact the immune response, priming substrates for glycosidase or protease degradation [210]. Moreover, NAs can improve binding of bacterial toxins, promote pro- inflammatory responses, and allow production of biofilms. NAs produced by the bacterial pathogen S. pneumoniae are critical for respiratory tract colonization. During infection, NA is involved in unmasking potential cell receptors [211-213] and in biofilm formation [184, 214]. S. pneumoniae produces three different enzymes that cleave sialic acid in various ways [201]. Only
NanA has the same enzymatic activity as the NA of influenza. NanA is extracellular, but remains attached to the bacterial surface. Only NanA has been demonstrated to play an important role during respiratory tract infection [215, 216]. Other NA-producing pathogens in the gastrointestinal, oral, or reproductive tract include Salmonellae, Vibrios, Clostridium spp., S. oralis, Porphrymonas gingivalis, Gardnerella vaginalis and Bacteroides [210].
Despite low amino acid similarity, all NAs share a common structure (Figure 2.1). The consensus structure of influenza NA is described as a β-propeller tetramer [124, 217]. Each monomer is formed by six four-stranded antiparallel β-sheets arranged in blades, providing a rigid catalytic site. NA produced by Salmonella typhimurium, which shares only 15% sequence similarity to the influenza NAs, also displays the clustered antiparallel β-sheets surrounding the active site [218]. Nonetheless, the length of the beta-strands and loops vary significantly. While displaying up to a 50% variation in the primary sequence, the active site residues of NA from
51 | P a g e
influenza A and B viruses are highly conserved and are similar to bacterial NAs. In the active site (Figure 2.1C), three arginine residues, a glutamic acid residue, and a tyrosine residue stabilize binding to the natural substrate, NeuAc [124, 217-220]. Calcium-binding sites have been reported. Up to three calcium-binding sites are present in each monomer of influenza NA
(Figure 2.1). One is a high affinity Ca2+ binding site, which is linked to NeuAc binding site [124].
The second site has low affinity for Ca2+ [221], and the third is located in the fourfold symmetry axis of NA tetramer [134] (Figure 2.1C). Interestingly, all of the amino acids responsible for binding calcium are located in loops near NeuAc binding residues (Figure 2.1A). Previous reports suggest that calcium contributes to the activity of influenza A neuraminidase [128, 129], but not the type B NA activity [129], affecting only the thermostability of type B NA [127].
Bacterial NAs, such as Vibrio cholerae sialidase, are reported to have up to 4 Ca2+ binding sites
[222], while NanA of S. pneumoniae has none [219].
Given the central role of NA in pathogenesis, host range restriction, antigenic variation, and as a therapeutic target, we investigated the enzymatic properties of NA from four different human respiratory pathogens, influenza A and B, parainfluenza and S. pneumoniae using the fluorogenic substrate, 2′-(4-methylumbelliferyl)-α- D -N- acetylneuraminic acid (MUNANA). The primary focus of our analysis was on whole viral particles and bacteria, as intact microorganisms are the target of antimicrobial agents. However, in some cases the properties of purified recombinant NA were examined and found to be very different from the activities observed on intact virus particles. NAs with different NAI susceptibility were assessed at various pH and calcium concentrations in the presence or absence of NAI. While pathogen-specific differences in activity were observed as a function of pH values, calcium dependence was only observed at suboptimal pH.
52 | P a g e
Figure 2.1. Diagrams of NA sequence and structure. A. Schematic of the influenza NA sequence. NA is oriented with its N-terminus in the virus membrane. A highly conserved signal sequence of polar amino acids at the N-terminus of the NA is followed by a transmembrane region formed by a sequence of hydrophobic amino acids. The signal peptide is not removed and no processing at the C-terminus takes place. NeuAc binding residues are indicated at the bottom of the diagram. The positions of mutations that confer NAI-resistance are shown as dotted grey lines. These include N1 mutations (I222V/R, H274Y); N2 mutations (E119V/I, I222V, R292K, N294S); and type B NA mutations (R152K, D198N/E I222T, N294S, and R371K). Calcium binding residues are shown as black circles at the top of the cartoon. The size of the circle correlates with the number of residues. B. Distance tree of viral NA amino acid sequences. This tree was produced using BLAST pairwise alignments. Neighbor joining algorithm was used, maximum sequence difference is 0.85 and Grishin distance is shown as a black arrow. S. pneumoniae NA sequence have a fraction of mismatched bases for any pair of sequences which is larger than maximum sequence difference; therefore, it was excluded from tree generation. C. Crystal structure of NA from different phylogenetic sources. The crystal structures of β-propeller monomer are shown for all NAs as light green cartoons. Calcium ions are shown as blue spheres. Disulfide bonds are shown as red sticks. Conserved NeuAc binding residues – three arginine residues, a tyrosine, and a glutamic acid – are shown in yellow. PyMOL software and Protein Data Base (PDB) files (3NSS, influenza N1; 2BAT, influenza N2; 1NSB, influenza type B NA; 1V3C, parainfluenza HN; and 2YA5, S. pneumoniae NanA) were used to generate the figures.
3 Experimental procedures
3.1 Viruses, cells and recombinant NA
Influenza viruses were originally obtained from the Centers for Disease Control (CDC).
Detailed list of virus used in this study is described in Table 2.1. The viruses were propagated in
53 | P a g e
Madin-Darby canine kidney (MDCK) cells (Diagnostic Hybrids Headquarters, OH) and were harvested after cytopathogenic effects were observed post virus infection. Infected tissue culture cells were centrifuged for 10 min at 4°C and stored at –80°C. Mock infected MDCK cells were used as a negative control.
The recombinant influenza N1 proteins (Table 2.1) were obtained from Sino Biological Inc.
(Beijing, P.R. China). Parainfluenza virus type 2 was obtained from Biodefense and Emerging
Infections Research Resources Repository (BEI number NR-3229). S. pneumoniae strains
ATCC 6305 (Strep 6) and ATCC 49619 (Strep 4) were kindly provided by Vicki Stegner at
University of Cincinnati Health Laboratory.
54 | P a g e
Table 2.1. Summary of characteristics of virus strains, recombinant NA protein, and S. pneumoniae strains evaluated in this study
% of Sequence comparisong e f 2+ NA FFU max NA Osel Zan Ca binding residues i HA Osel GenBank Virus strain a act. per act. Ki Ki Other Ct b Titer c 293 294 297 298 324 326344 348 Resistance # RFU mL +EDTA (nM) (nM) DN GS DNP NGANG h polymorphism d mutation (pH) H1N1 strains A/California/07/09 44% Used as query 23 6600 320 7 x 108 0.9 1.3 ▪▪j ▪▪ ▪▪▪ ▪▪▪▪▪ - ACQ63272 (H1N1pdm/09-S) (pH 6-7) sequence A/North Carolina/39/09 63% 105V→I, 25 5300 80 5 x 106 125 1.2 ▪▪ ▪▪ ▪▪▪ ▪▪▪▪▪ 274H→Y EPI233169k ( H1N1pdm/09-R) (pH 6-7) 246N→D A/Brisbane/59/07-Like 54% Used as query 25 5000 160 3 x 107 2220 3.3 ▪ ▪ ▪▪ ▪▪▪ D ▪ ▪ ▪ ▪ 274H→Y ADE28752 (H1N1/07-R) (pH 6) sequence A/Florida/21/08 36% 263V→I, 25 4700 320 2 x 106 1729 2.2 ▪▪ ▪▪ ▪▪▪ ▪▪▪▪▪ 274H→Y ACM51313 (H1N1/08-R) (pH 6) 351D→G Recombinant N1 protein Identical to A/California/04/09 94% n/a 16000 n/al n/a 62 25 ▪▪ ▪▪ ▪▪▪ ▪▪▪▪▪ - H1N1pdm/09- ACQ63272 (N1pdm/09-S) (pH 6-8) S Identical to A/California/04/09 94% ACQ63272 n/a 13000 n/a n/a 758 3.7 ▪▪ ▪▪ ▪▪▪ ▪▪▪▪▪ 274H→Y H1N1pdm/09- (N1pdm/09-R) (pH 6-8) 274H→Y R >25 far from A/Anhui/01/05 43% n/a 12900 n/a n/a 13 5.6 ▪▪ ▪▪ ▪▪▪ ▪▪▪Y▪ - Ca2+ and ABU94738 (N1/05-S) (pH 6-7) NeuAc sites H3N2 strains A/Brisbane/10/07-Like 6 50% Used as query 23 6700 40 3 x 10 0.6 1.6 ▪▪ ▪▪ ▪T▪ E▪ GH▪ - ACO95273 (H3N2/07-S) (pH 7) sequence 9T→A, 26I→T, A/Texas/12/07 61% 25 6800 160 4 x 106 15 2.6 ▪▪ ▪▪ ▪T▪ E▪ GH▪ 119E→V 312T→I, ACA33536 (H3N2/07-R) (pH 7) 435E→V Type B strains B/Brisbane/60/08 7 92% Used as query 28 5800 320 5 x 10 13 4.1 ▪▪ TA ▪T▪ K▪ SG▪ - ACN29381 (Vict/06-S) (pH 7) sequence >15 far from B/Memphis/20/96 25% 25 5900 320 9 x 103 104 6.3 ▪▪ TA ▪T▪ K▪ SG▪ 150R→K Ca2+ and ACT85963 (Vict/96-R) (pH 7) NeuAc sites
55 | P a g e
>20 far from B/Florida/04/06 88% 27 5000 320 9 x 105 19 4.5 ▪▪ TA ▪T▪ K▪ SG▪ - Ca2+ and ACA33351 (Yama/08-S) (pH 7) NeuAc sites Parainfluenza virus Type m 27% n/a 4800 ND ND ND ND n/a n/a n/a n/a n/a n/a AEQ39010 2 (PIV) (pH 5) S. pneumoniae 112% n/a 5000 n/a n/a ND ND n/a n/a n/a n/a n/a n/a CAI94558 ATCC 6305 (Strep 6) (pH 6-9) S. pneumoniae 108% n/a 5000 n/a n/a ND ND n/a n/a n/a n/a n/a n/a CAI94558 ATCC 49619 (Strep 4) (pH 6-9) a: Rounded threshold cycle (Ct) values determined by RT-PCR in viral samples at 1 to 1000 dilution. b: Rounded average of RFU/hour/μL in NA assay in PBS, pH 7 for all influenza and S. pneumoniae strains, and sodium acetate buffer, pH 5 for Parainfluenza type 2. c: Fluorescent Focus units were determined using the fluorescent focus-forming assay d: % of maximal NA activity obtained in presence of 25 mM EDTA. All influenza strains tested at pH 7 (Figure 3G). Parainfluenza and S. pneumoniae tested at pH 5 and pH 7, respectively (Figure 4) e: Osel: Oseltamivir f: Zan: Zanamivir g: Sequence comparison using BLAST sequence analysis tool at National Center for Biotechnology Information (www.ncbi.nlm.nih.gov) using the indicated pairs h: NAI resistance mutation i: GenBank accession number at National Center for Biotechnology Information j: “▪” is used for nucleotide identities k: Gene accession number at Global Initiative on Sharing Avian Influenza Data (GISAID, http://platform.gisaid.org) l: n/a: Not applicable m: ND: Not determined
56 | P a g e
3.2 Hemagglutinin (HA) titer
HA titration of influenza isolates grown in cell culture was performed as previously described [68]. Briefly, serial 2-fold dilutions of influenza virus in phosphate buffered saline
(PBS) buffer (8.1 mM Na2HPO4, 1.5 mM KH2PO4, 128 mM NaCl, 2.7 mM KCl; pH 7.4) were incubated with a standardized turkey red blood cell (RBC; ViroMed, Minnetonka, MN) suspension (0.5% erythrocytes) in the wells of a microtiter plate. The plates were mixed and the
RBCs were allowed to settle for 30 minutes at room temperature. The plate was tilted and read for complete agglutination. The inverse of the last dilution of virus that causes hemagglutination is considered the HA titer.
3.3 Influenza Fluorescent Focus forming Assay (FFA)
The influenza FFA was used to detect and quantify infectious influenza particles in cells
[68]. Briefly, MDCK cells were grown in 96-well plates to confluency. Serially diluted virus preparations were added to the cells and incubated for 2 hours at room temperature. The infected cultures were washed, overlaid with media, and incubated for 20 h at 33°C, 5% CO2.
Cells were fixed with acetone, and virus was detected using monoclonal antibody against influenza A or influenza B nucleoproteins (Millipore, USA) along with fluorescent secondary conjugated antibody (Millipore, USA). The fixed cell monolayers were viewed using UV light microscopy and influenza positive cells, visible as fluorescent foci, were counted. The titer was calculated by multiplying the number of foci by the inverse of the dilution and dividing by the volume of inoculum. Final fluorescent focus forming units (FFU) were determined using the average of duplicate wells.
3.4 Real time Reverse Transcription-Polymerase Chain Reaction (RT-PCR)
Real time RT-PCR procedures recommended by the World Health Organization (WHO)
[223] were used to detect viral genomes. Briefly, viral RNA was isolated from 140 μl of samples using a QIAamp® Viral RNA Kit (Qiagen, Germany) and the cDNA synthesis was carried using an Omniscript TM kit (Qiagen, Germany) according to the manufacturer’s instructions. In vitro
57 | P a g e
cultured viral samples were diluted 1000-fold, and a total of 5 μl of the diluted RNA was used in the PCR reaction which was amplified using a mixture of primers specific to highly conserved regions of the influenza genome. Final concentration of primers was 40 µM and concentration of probes was 10 µM in each PCR reaction mixture. Ambion AgPath-ID™ One-Step RT-PCR Kit
(Ambion, USA) was used in the PCR reaction carried out according to the manufacturer’s instructions under the following conditions: initial holds at 50 ºC for 30 min and 95 ºC for 10 min followed by 45 cycles 95 ºC for 15 sec and 55 ºC for 34 sec. Conditions for data collection were
55 °C for 34 min. Threshold cycle (Ct) values are considered as the number of cycles required for the fluorescent signal to cross the threshold or background level. All primers and probes were obtained from Biosearch Technologies Inc., USA.
Influenza H1N1pdm A nucleocapsid protein (NP), used for A/California/07/09
(H1N1pdm/09-S) and A/North Carolina/39/09 (H1N1pdm/09-R): Forward primer 5'-
GCACGGTCAGCACTTATYCTRAG-3', Reverse primer 5'-GTGRGCTGGGTTTTCATTTGGTC-
3', Probe 6-carboxyfluorescein (6-FAM)- 5'-CYACTGCAAGCCCATACACACAAGCAGGCA-3' labeled internally black hole quencher-1 (BHQ-1).
Influenza A matrix protein (M1), used for all other influenza A isolates: Forward primer 5'-
GACCRATCCTGTCACCTCTGAC-3', Reverse primer 5'-AGGGCATTYTGGACAAAKCGTCTA-
3', Probe 6-FAM-5'-TGCAGTCCTCGCTCACTGGGCACG-3'-BHQ-1.
Non-structural 1 (NS1) gene influenza B, used for all influenza B isolates: Forward primer
5'-TCCTCAAYTCACTCTTCGAGCG-3', Reverse primer 5'-CGGTGCTCTTGACCAAATTGG-3',
Probe 6-FAM-5'-CCAATTCGAGCAGCTGAAACTGCGGTG-3'-BHQ-1
3.5 NA pH dependence
NA activity was examined using the fluorescent substrate, MUNANA (Sigma-Aldrich, Inc.,
USA). Viruses were diluted 50-fold in five different reaction buffers, all containing 10 µM
MUNANA, 200 mM NaCl and 0.1 mM calcium. Sodium acetate buffer (50 mM) was used for studies at pH 4, 5 and 6. PBS was used for studies at pH 6, 7 and 8. Tris(hydroxymethyl)
58 | P a g e
aminomethane (Tris) buffer (50 mM) was used for studies at pH 8 and 9. N-cyclohexyl-2- aminoethanesulfonic acid (CHES) buffer (50 mM) was used for studies at pH 9 and 10. 3-
(cyclohexylamino)-1-propanesulfonic acid (CAPS) buffer (50 mM) was used for studies at pH 10 and 11. Similarly, colonies of S. pneumoniae were suspended to an optical density of 1 at 600 nm in water and 50-fold dilutions were made in each of the reaction buffers. All reactions were incubated at 37 °C for 3 h. The pH optimum for detecting fluorescence was determined to be pH
8-9. All samples were adjusted to pH 8 using 1 M Tris buffer before release of fluorescent 4- methylumbelliferone was measured with FLX-800 fluorimeter (BioTek, USA) using excitation and emission wavelengths of 350 and 460 nm, respectively. Relative fluorescent units (RFU) were used to quantify NA activity. The results are collected from at least three independent experiments.
3.6 NA calcium dependence
Both influenza and parainfluenza viruses were diluted 20-fold in HEPES buffer (50 mM
HEPES, 200 mM NaCl; pH 7 for influenza virus and pH 6 for parainfluenza virus) containing 10
µM MUNANA, with different concentrations of CaCl2, and 0 mM of CaCl2 was designated when
25 mM ethylenediaminetetraacetic acid (EDTA) was added. Incubation time and fluorescence measurement conditions were examined as described in pH dependence assay in this study.
The results were collected from at least three independent experiments. RFUs were used to quantify NA activity and percentages to NA activity were calculated, defining 100% as the largest value in each data set, using Prism 5.0 (GraphPad Software, USA) for the analysis.
3.7 Limit of detection of NA assay
The limit of detection of influenza virus was determined by measuring NA activity in different dilutions of viral samples in PBS containing 10 μM MUNANA and 0.1 mM calcium at pH
6.5, while pH 7 was used to test S. pneumoniae NA activity. Different dilutions of virus or bacteria were incubated at 37°C for 3 hours and read as described above. The results are compiled from at least three independent experiments in each dilution.
59 | P a g e
3.8 NAI susceptibility assay
Drug susceptibility testing was performed by the 50% inhibitory concentration method, using two different procedures. First, we co-incubated the virus with the fluorogenic substrate
MUNANA and NAI. Second, we pre-incubated the virus with NAI at 37°C for 30 min, followed by the addition MUNANA. NA activity was measured in the presence of increasing concentrations of active oseltamivir (oseltamivir carboxylate, Toronto Research Chemicals Inc., Canada) and zanamivir (GlaxoSmithKline, USA). All experiments were performed at the optimal pH of each strain with 0.1 mM of CaCl2. The inhibition constant (Ki) was calculated using Prism 5.0
(GraphPad Software, USA). The results are compiled from at least three independent experiments.
4 Results
In this study, six influenza A strains, three influenza B strains, one parainfluenza strain, and two S. pneumoniae strains were studied (Table 2.1). When possible, epidemiologically significant strains were chosen. Attempts were also made to use closely related oseltamivir- resistant influenza strains (indicated by –R) and oseltamivir-sensitive influenza strains (indicated by –S) (Figure 2.1B). Viral titers were quantified by four different parameters; number of viral genomes (PCR Ct value), amount of functional surface proteins (HA titer and NA activity), and infectivity (FFU). All of the influenza preparations possessed similar levels of NA activity
(ranging from about 4700 to 6800 RFU/hour/µL) and comparable Ct values (ranging from 23 to
28) showing that abundant material was present in the viral preparations [224]. However, HA titers varied from 40 to 320, and FFU varied from about 104 to almost 109, and there was no obvious correlation between HA and FFU titers. In summary, the virus samples have similar NA activity and Ct values, suggesting similar levels of virus nucleic acid and NA protein are present.
However, their biological activities including host receptor recognition (HA titers) and infectivity
60 | P a g e
(FFU values) differ greatly. Previous studies have also indicated that influenza virus samples vary significantly in HA titers and infectivity units [225].
4.1 Characterization of the optimal pH for NA activity
The influence of pH on NA activity was determined in the presence of 0.1 mM calcium
(Figure 2.2). All influenza strains displayed significant NA activity between pH 6 to 7.
Interestingly, N1 viruses retained greater than 75% of the NA activity at pH 5. All recombinant
N1 proteins assayed, N1pdm/09-S, N1pdm/09-R and N1/05-S showed activity at pH 5 (Figure
2.2, second row). H3N2 and type B Victoria influenza strains (Figure 2.2, third and fourth row), displayed little detectable activity at pH 5. Interestingly, parainfluenza type 2 virus NA activity is optimal at pH 5, while at pH 7 the enzymatic activity is reduced to below 25% of maximal activity. These values which agree with the reported optimum (pH 5) for parainfluenza type 3
[200]. Contrary to viral NA activity, the S. pneumoniae NA displayed a pH optimum of 6-9
(Figure 2.2, last row).
61 | P a g e
Figure 2.2. Influence of pH on NA activity. Results are grouped by NA strain. NA activity was measured by incubating virus with fluorogenic substrate at different pH values. NA activity in acetate buffer pH 4, 5 and 6 is shown as closed squares (■), in PBS pH 6, 7 and 8 is shown as open circles (O), in Tris buffer pH 8 and 9 is shown as closed diamonds (♦), in CHES buffer pH 9 and 10 is shown as open triangles (∆), in CAPS buffer pH 10 and 11 is shown as asterisk (*). The plots show results of at least 3 independent experiments and error bars indicate standard deviation (SD).
4.2 Characterization of the role of calcium in NA activity
In studies performed several decades ago, divalent cations, including Ca2+, Mn2+, Mg2+, and Fe2+ were reported to stimulate N1 activity of up to 100-fold for both purified protein and whole viral particles [128, 129], with Ca2+ having the maximal effect. Similar results were reported for 18 avian NAs (subtypes N1, N2, N3, N6, and N7), but not with human type B NA
[129], which has not been reported to have Ca2+-dependent activity. We examined the influence of calcium on NA activity for recently circulating influenza virus, including oseltamivir -sensitive and -resistant strains of H1N1, H3N2, and influenza type B.
NA activity of whole viral particles and recombinant NA was tested at different calcium concentrations including the absence of calcium, ensured by the addition of the chelator EDTA at 25 mM (Figure 2.3). For all influenza and parainfluenza viruses tested, nearly maximal
62 | P a g e
activity was observed at calcium concentration between 0.09 mM to 100 mM. Interestingly, removal of calcium by the addition of 25 mM EDTA had very little effect on NA activity except
Vict/96-R (Figure 2.3E and 2.3G) and purified recombinant protein N1/05-S (Figure 2.3C and
2.3G). Very high calcium levels (500 mM) were inhibitory for all viruses; however, purified recombinant proteins, N1pdm/09-S and N1pdm/09-R displayed significant activity in the presence of 500 mM calcium (Figure 2.3C)
Comparison of enzymatic activity for oseltamivir-sensitive (Figure 2.3G, black bars) and oseltamivir-resistant strains (Figure 2.3G, open bars) in the presence of EDTA chelator is shown. The H1N1pdm/09 oseltamivir-sensitive strain displays significantly less activity in the presence of EDTA than the closely related resistant strain, while no difference was seen when the corresponding purified recombinant proteins were examined. The NA activity of closely related H3N2 oseltamivir-sensitive and -resistant strains were inhibited to a similar degree, while activity of the oseltamivir-resistant influenza B strain Vict/96-R was significantly reduced compared to Vict/06-S in the presence of EDTA.
63 | P a g e
Figure 2.3. Influence calcium on NA activity. A, H1N1pdm/09 influenza type A H1N1 strains; B, human influenza type A H1N1; C, recombinant influenza N1 proteins; D, influenza type A H3N2; E, influenza type B from Victoria and Yamagata linages; F, parainfluenza virus type 2; and G, Statistical analysis of NA activity was evaluated in the presence of 25 mM EDTA chelator and compared to the NA activity when 0.1 mM calcium was present. Assays were performed at pH 7 for the influenza viruses, and parainfluenza virus. Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5. The plots show results of at least 3 independent experiments and the error bars indicate SD.
4.3 NAs from different phylogenetic sources display unique calcium and pH profiles.
We examined NA activity in the presence and absence of calcium at pH 4, 5 and 7
(Figure 2.4). Interestingly NA within each phylogenetic group displayed a similar profile. The influenza N1 and N2 groups are most clearly distinguished by the robust activity at pH 5 in the presence of calcium for the N1 viruses, and lack of activity at pH 5 for the N2 viruses. Similarly, for the influenza B viruses, Yamagata was similar to influenza A, N1, while influenza B Victoria
64 | P a g e
was similar to influenza A, N2. Parainfluenza is distinguished by significant calcium-dependent
NA activity at pH 4 and 5. S. pneumoniae is distinguished by high NA activity and calcium- independent activity at pH 5 and 7 (Figure 2.4). Influenza can also be differentiated from S. pneumoniae by a little or no influenza NA activity at pH 9, while S. pneumoniae display a high
NA activity at pH 9 (Figure 2.2).
Figure 2.4. Comparison of NA from different phylogenetic groups. Strains are grouped according to similarity in patterns of NA activity measured at different pH and calcium conditions. Calcium (+) indicates NA activity measured in the presence of 0.1 mM Ca2+; calcium (-) indicates NA activity measures in the presence of 25 mM of EDTA chelator. The plots show results of at least 3 independent experiments and the error bars indicate SD.
4.4 Limit of detection of NA assay
To determine the limit of detection of influenza virus, we measured the NA activity of serial dilutions of virus. In all cases, we detected viral strains down to 500-fold dilution (Figure 2.5), which correspond to range of 10 to 106 FFU (Table 2.1). We also determined the limit of detection of both S. pneumoniae strains in PBS to be a 45000-fold dilution (Figure 2.5E).
65 | P a g e
Figure 2.5. Limit of detection of influenza and S. pneumoniae NA activity. A, Strains are grouped according to similarity. NA was measured by incubating dilutions of virus or bacteria with MUNANA substrate diluted in PBS, pH 6.5 with 0.1 mM Ca2+. A, H1N1pdm/09 influenza type A H1N1 strains; B, human influenza type A H1N1; C, influenza type A H3N2; D, influenza type B from Victoria and Yamagata linages; E, S. pneumoniae strains. The plots show results of at least 3 independent experiments and the error bars indicate SD Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5. NA activity of all dilutions of each influenza and S. pneumoniae strains was compared individually to the negative control. The last dilution in which there was a statistically significant difference from the negative control is indicated by *** with a P value smaller than 0.0001 in each case.
4.5 NAI susceptibility assay
Two different approaches are commonly used to determine the Ki. In the pre-incubation approach, the enzyme is first incubated with the inhibitor, and then substrate is added. In the co- incubation approach, the inhibitor and substrate are added at the same time. Oseltamivir is a prodrug that is metabolized to the active form, oseltamivir carboxylate, in the liver. The inhibition constant (Ki) was determined for oseltamivir carboxylate using both approaches (Figure 2.6), and representative data demonstrates that the two approaches give very similar Ki values.
The Ki values determined using the co-incubation method for both oseltamivir carboxylate and zanamivir are summarized in Table 2.1. All viruses were sensitive to concentrations of zanamivir ranging from 1 to 10 nM. As seen in the pH and calcium studies, significant
66 | P a g e
differences in oseltamivir carboxylate Ki were seen for viral particles and purified recombinant protein reported to have identical NA sequences. For example, for oseltamivir-sensitive viral particle H1N1pdm/09 (Ki = 0.9 nM) and its corresponding recombinant NA N1pdm/09-S (Ki = 62 nM) differ by 60-fold, and 19-fold for zanamivir. Less dramatic differences were seen for Ki determined viral particle H1N1pdm/09-R and its corresponding recombinant NA N1pd/09-R, which differs by 6-fold for oseltamivir susceptibility and 3-fold for zanamivir susceptibility.
When viral types are compared, type A virus (H1N1pdm/09-S, Ki = 0.9 nM and H3N2/07-S;
Ki = 0.6 nM) are about 10-100 fold more sensitive to oseltamivir than type B influenza strains
(Vict/06-S, Ki = 13 nM and Yam/08-S, Ki = 104 nM), despite type B influenza strains being reported to be sensitive to oseltamivir. The His274Tyr mutation is commonly associated with oseltamivir-resistance in influenza A N1. Interestingly, dramatically different levels of resistance are seen in type A H1N1 strains with this resistance mutation. For example, the Kis for the
H1N1/07-R (Ki = 2220 nM) and H1N1/08-R (Ki = 1729 nM) strains are more than 10-fold greater than the H1N1pdm/09-R strain (Ki = 125 nM). The resistance level of H3N2/07-R with the
Glu119Val mutation (Ki = 15 nM) is 10-100 times lower than the H1N1 strains with His274Tyr mutation. Surprisingly, the Ki of oseltamivir-sensitive Yama/08-S (Ki = 19), which lacks resistance mutations, is similar to the oseltamivir-resistant H3N2/07-R strain (Ki = 15 nM), which contains a resistance mutation.
67 | P a g e
Figure 2.6. Evaluation of co-incubation and pre-incubation of oseltamivir on NA activity. NA activity was measured using two different procedures as detailed in the material and methods. The top panel shows co-incubation of different concentrations of oseltamivir with NeuAc substrate. The bottom panel shows pre-incubation of different concentrations of oseltamivir with NeuAc substrate. Ki values were determined using Prism 5.0 (GraphPad Software, La Jolla, CA).
5 Discussion
The present study characterized NA activity for different phylogenetic groups. The HA and
NA activities of influenza and parainfluenza viruses both recognize sialic acid, but HA and NA perform different functions in the viral life cycle. During influenza infection, HA activity binds to sialic acid to promote cellular entry. In contrast, NA activity cleaves sialic acid from host and viral proteins and promotes viral release after the infectious cycle is completed. Since inappropriate
NA activity could remove the sialic acid receptor from uninfected host cells and inhibit viral attachment, the two activities must be separated in space or time. In influenza, the HA and NA activities reside on separate proteins, and are spatially segregated on the virion. More HA is present on the influenza viral particle than NA, and the two proteins are localized in separate clusters. This arrangement is thought to reduce access of NA to potential cellular receptors engaged by HA. In contrast, parainfluenza produces a single protein with both HA and NA activity, called hemagglutinin–neuraminidase or HN [190, 203]. The HA and NA activities reside on the same molecule, but display a temporal segregation based on the viral life cycle. At pH 7, the pH of the healthy human respiratory tract [226, 227], parainfluenza NA activity is weak.
However, it is highly active at pH 5 (Figure 2.2), and necrosis following cellular death of infected cells could create the acidic environment favored by parainfluenza NA. Thus, the preference of
68 | P a g e
the parainfluenza NA for low pH would result in activity associated with lysed cells and little activity in the presence of healthy tissues. Thus, sialic acid in the host membrane would not be removed from healthy cells, ensuring the presence of receptor for parainfluenza HA.
While NA from some influenza strains have been reported to be active at low pH, this property was primarily attributed to avian viruses, which are adapted for replication in digestive tract of hosts [108, 228]. Moreover, residues Thr435, Arg430 and Gly455 in N1 [229] of influenza NA were identified to be involved in low-pH enzymatic activity. Deletion of Thr435 and mutation Asn455Gly stabilize NA activity at low pH [229]. However recently, the low-pH activity of both human N1 and N2 influenza viruses has been reported to contribute to virus replication efficiency in vitro [14, 229]. All the H1N1 viral particles along with human N1 recombinant protein studied here contain both Thr435 and Gly455 residues and show activity at low pH 5 (Figure
2.2). Taken together, the ability of NA to be active at low pH is suggested to be a characteristic that facilitates viral pathogenicity and transmission in vivo and in vitro, not an exclusive characteristic of avian influenza virus [14, 230].
In our studies, high S. pneumoniae NA activity (>75% of maximal NA activity) was observed to occur in a broad range of pH 5 to 9 (Figure 2.2). NA activity of S. pneumoniae is critical for respiratory tract colonization and the pH range of NA activity observed here correlates with the lower respiratory tract pH, ranging from 6.4 to 7.7 [226] and the reported enzymatic activity of purified NanA neuraminidase, which occur in a pH rage of 5 to 8 having its optimal at
7 [202]. We did not observe calcium dependence (Figure 2.4), consistent with the reported lack of calcium binding sites in NanA.
A total of three Ca2+-binding sites have been reported for influenza NA monomers [123,
124, 131-134] (Figure 2.1). All N1, N2 and type B NA have one high-affinity calcium binding site located near NeuAc binding site. A second Ca2+-binding site with low-affinity was reported for N1 and influenza B NA (Figure 2.1C), which is located at molecular fourfold axis of the NA tetramer. A third Ca2+-binding site with low-affinity was reported for N1 and it is located far from
69 | P a g e
substrate site (Figure 2.1C). Our results show that lack of calcium typically results in only about a 50% decrease in activity when assayed at the optimal pH. These results differ from older reports where egg-grown purified influenza virus, NA protein fractions, and indirect methods to measure NA activity were used [127, 128]. The difference might be explained by the fact that we tested NA activity of MDCK-grown whole influenza viral particles and directly measured activity using a fluorescent substrate. Our assay appears to better represent physiological conditions, and the modest influence of calcium depletion on NA enzymatic activity suggests that Ca2+ is not needed to maintain the enzymatically active conformation. All of the calcium binding sites are located in loops which connect the β-sheets of NA, and the residues that bind calcium are near to NeuAc binding residues (Figure 2.1). Calcium likely plays a role in folding of β-propeller tetramer. The structure of the mature NA is stabilized by disulfide bonds [231, 232] and it is possible that calcium promotes correct folding initially, but is not critical for activity after the disulfide bonds are formed. Similar mechanisms have been reported for other proteins such as calmodulin [233], Clostridium spp. dockerin [234], and Pseudomonas spp. alkaline protease
[235], where correct folding depends on Ca2+.
Previous reports based on X-ray crystallography [136] and molecular modeling studies
[127, 137] of influenza NA have suggested that the absence of calcium will result in a greater loss of activity of NAI resistant NA than NAI sensitive NA [127, 136, 137]. This was not observed in type A influenza H1N1 and H3N2 viral particle, or N1 recombinant protein (Figure 2.3G). Only oseltamivir-resistant type B influenza (Vict/96-R, Figure 2.3E) displayed loss of activity in the absence of calcium. Our results demonstrate that oseltamivir-resistant influenza virus type A neuraminidases are not more susceptible to the loss of enzymatic activity in the absence of calcium than oseltamivir-sensitive NA.
With the emergence of bacterial strains resistant to all available antibiotics, the medical community has become painfully aware of the importance of antimicrobial stewardship.
Administration of antimicrobial agents in absence of clinical benefit is the major driving force in
70 | P a g e
evolution of antimicrobial resistance. With regard to influenza, treatment of patients infected with partially NAI resistant isolates will encourage the selection and amplification of mutations conferring even higher levels of resistance. In our NAI studies, we found that resistance levels of oseltamivir-resistant type A influenza viruses with His274Tyr mutation can differ by 20-fold
(Table 2.1). Interestingly, the bioavailable concentration of oseltamivir in treated patients is about 350 nM [236]. This concentration will inhibit 70% of NA activity of oseltamivir-resistant
H1N1pdm/09-R with His274Tyr (Table 2.1), while not significantly affecting other oseltamivir- resistant strains (H1N1/07-R and H1N1/08-R) (Table 2.1). Moreover, our results show that the resistance level of oseltamivir-resistant H3N2/07-R with the Glu119Val mutation is equal to oseltamivir-sensitive Yama/08-S and it is up to 100 times lower than the H1N1-R strains with
His274Tyr. NAI sensitive or resistant strains are often identified based on the presence of point mutations such as His274Tyr in N1, Glu119Val in N2 or Lys150Arg in Type B NA by PCR and pyrosequencing [81, 82, 237-240]. Since the same point mutations on NA can confer different levels of resistance to NAI (Table 2.1), the differentiation of NAI- resistant from NAI- sensitive strains should be based on the level the inhibitory concentration of the drug rather than the presence of point mutations on the NA sequence. Understanding the true susceptibility to NAIs is essential for implementing effective diagnostic and treatment strategies for influenza.
We also observed that viral particles are more susceptible to both oseltamivir and zanamivir than purified recombinant NA proteins (Table 2.1). In viral particles, NA is found as a tetramer while the recombinant NA is found in different oligomeric forms (e.g. dimer and monomers). Wu et al. reported that differences in the N-glycosylation patterns of NA tetramers versus NA monomers or dimers influence oligomerization, and thus NA enzymatic activity [241].
NA monomers and dimers show an unfavorable conformation with low affinity for NeuAc substrate. Only NA tetramers had a fully active conformation. Our results suggest that NAI susceptibility studies should be performed with intact virus, not purified recombinant protein.
71 | P a g e
In summary, determination of NA activity at different pH and calcium conditions can differentiate human respiratory pathogens (Figure 2.4). Assays performed in the presence and absence of NAI can direct the appropriate use of antiviral therapy. Previously, the usage of rapid diagnosis of influenza by pediatric departments is reported to reduce the number of radiographs ordered, decrease the use of antibiotics and decrease length of time to discharge increasing antiviral therapy use [242, 243]. The approach described here shows promise for the development of rapid, sensitive, and highly specific point-of-care diagnostics.
The results of this chapter can be found in:
Karen M. Gallegos, Nicole R. Meyer, Monica M. McNeal, and Alison A. Weiss. (2013) Comparison of NA activity from influenza, parainfluenza and Streptococcus pneumoniae. Manuscript submitted for publication.
72 | P a g e
CHAPTER 3: STUDIES ON INFLUENZA BINDING
1 Abstract
Influenza infections result in half million deaths per year worldwide. Rapid diagnosis of influenza could reduce disease burden; however, antigenic drift and shift limit the sensitivity of traditional antibody-based tests. We analyzed the ability of host factors to bind influenza surface proteins to potentially replace antibodies as capturing agents. Three types of host factors were considered: (1) factors that contain sialic acid, the traditional influenza receptor (e.g. synthetic sialylated glycans or the bovine glycoprotein, fetuin); (2) factors that bind to the high mannose core present on viral proteins (e.g. Dendritic Cell Specific ICAM3 Grabbing Nonintegrin (DC-
SIGN) analogs, mannose binding lectin (MBL) and macrophage mannose receptor, MMR); and
(3) factors that play a role in the interaction of influenza with host cells (e.g. epidermal growth factor receptor (EGFR) and galectin). Binding of influenza virus types A and B to host factors was assessed by a capturing assay. Briefly, host factors were immobilized in microtiter plates and incubated with dilutions of influenza virus. Neuraminidase activity of bound virus was detected using a fluorogenic sialic acid substrate. In our capturing assay, we were able to detect binding of type A (H1N1 and H3N2) and type B (Victoria and Yamagata) influenza. Host proteins captured up to 20% of the input neuraminidase activity, with the overall rank order; fetuin > MBL
> MMR > DC-SIGN > EGFR > galectin. In contrast, synthetic glycans were about 10-fold less effective. Host factors directly bind to influenza virus and show promise for replacing antibodies in diagnostic applications.
73 | P a g e
2 Introduction
Up to 500,000 deaths per year worldwide are caused by influenza virus infections [1].
Influenza types A and B viruses are most responsible for causing disease in humans.
Considering the rapid course of viral action and severity of symptoms, early detection of
Influenza is critical. Influenza antigenic drift and shift rapidly changes the virus, making it challenging for detection. Many antibody-based influenza diagnostic tests show a deficient sensitivity for detection, particularly when new influenza strains emerge [244]. Molecular (e.g.
PCR-based tests) or cell culture based tests have high cost, require laboratory expertise, or the turnaround time is too slow to influence clinical decisions. For example, during the recent pandemic H1N1pdm/09, numerous false negative results in influenza diagnosis delayed treatment, thus potentially increasing infection severity and mortality [77-79]. Taken together, it is important to accurately detect new and circulating strains of influenza virus in order to improve influenza treatment and disease outcome.
As an alternative to antibody-based influenza diagnostic tests, we used a new approach to develop an assay for influenza detection based on invariant characteristics of the influenza virus. Influenza is a membrane virus with two antigenic surface glycoproteins: hemagglutinin
(HA) and neuraminidase (NA). HA is responsible for binding host receptors bearing terminal sialic acid (NeuAc) [10]; while NA is responsible for cleaving NeuAc present on host proteins or glycolipids. NA activity is essential for efficient replication; it increases access to the cell surface by removing decoy receptors on mucins, cilia, and the cellular glycocalyx [125, 126], and prevents self-aggregation of virus, guaranteeing the efficient spread of progeny virus [7]. HA binding to host receptors and NA enzymatic activity are properties of influenza that are retained even after antigenic drift and shift permitting detection of new, as well as, seasonal strains. We used NA enzymatic activity to detect bound virus, and assessed virus capture with three different types of host factors that interact with influenza during the course of human disease: binding to natural and synthetic molecules that display sialic acid, capture by host lectins that
74 | P a g e
recognize the unusual glycosylation patterns of viral proteins, and host proteins thought to be involved in virus internalization or adhesion.
It is well established that sialic acid, specifically, N-acetylneuraminic acid (NeuAc) found in glycoproteins and glycolipids is the host receptor for influenza virus [10]. The terminal NeuAc binds to residues at the globular head of the HA molecule, which are conserved throughout antigenic variation. While a few influenza viruses can bind well to NeuAc alone, some viruses require oligosaccharides with as many as five glycans of defined composition and linkage [104,
245]. A critical determinant is the linkage between NeuAc and the penultimate galactose residue naturally found in glycosylated proteins. Two forms predominate, NeuAcα2-3Gal and
NeuAcα2-6Gal. It was thought that human influenza A virus uses α2-6 glycans and avian influenza virus uses α2-3 glycans; however, more recent glycan array studies reveal that H1N1 viruses bound both NeuAcα2-6 and NeuAcα2-3 receptors [102, 103], and human respiratory epithelial cells express NeuAcα2-3 as well as NeuAcα2-6 glycans [99-101]. We will use these glycan binding properties of influenza HA to develop and assay to detect virus.
In addition to viral recognition of the host, some host proteins recognize the virus as part of the immune response to infection [157]. The glycosylation patterns on the influenza surface proteins HA and NA differ from host glycoproteins due to differences in trafficking patterns used to reach the cell surface. N-glycosylation for all proteins begins in the endoplasmic reticulum
(ER) membrane, where fourteen sugars are added to an Asn displayed in Asn-X-Ser/Thr motifs
(Figure 3.1, step A). Subsequently, remodeling of the N-glycan chain of the protein occurs in the ER and Golgi apparatus on a stepwise manner. The main mannose trimming occurs in the
Golgi apparatus (Figure 3.1 step C and E) which leads to the typical N-glycosylation pattern found on mature mammalian glycoproteins [246]. In contrast, viral proteins are glycosylated in the ER leading to a high mannose core (Figure 3.1, step A and B) [153], but appear to escape from the ER before further modification of the high mannose core occurs (Figure 3.1). Although the exact mechanism of viral protein escape is still not clear yet, coatomer (COP), which coats
75 | P a g e
membrane-bound transport vesicles, is reported to play a role in the escape of influenza proteins from the ER (Figure 3.1) [247].
Figure 3.1. The N-glycosylation process. A. N-glycosylation process is initiated in the ER by the addition of a 14-sugar conserved core to Asn residues. B. Glucose residues are removed, leading to high mannose core, and the protein is then transported to the Golgi apparatus. C Major trimmings of mannose occur in the cis-cisterna of Golgi apparatus and a GlcNAc is attached to the core. D. An additional GlcNAc is attached to the core yielding the precursor for all biantennary glycans. E. In the last step of the process, various galactose and sialic acid sugars are attached to the modified core leading to the typical N-glycosylation pattern in mammals. Figure generated using Chemdraw.
Several mammalian lectins recognize the high mannose pattern on viral glycoprotein [113,
155, 156] which include Dendritic Cell-Specific Intercellular Adhesion Molecule Grabbing Non-
Integrin (DC-SIGN) and its analogs [168-171]. DC-SIGN is a transmembrane C-type lectin, which requires calcium for binding, mediating signaling upon ligand binding through a cytoplasmic tail [159, 160]. Human DC-SIGN and DC-SIGN receptor (DC-SIGNR) have similar binding patterns, recognizing internal and terminal mannoses, fucosylated glycans, and certain
N-aceltylglucosamine (GlcNAc) residues [169, 248, 249]. Recognition by mouse DC-SIGN is more specific; since it has low affinity for individual mannose residues, but recognizes dense arrays of glycans more typical of high mannose residues [250-252]. Other mannose binding receptors which interact with influenza included Macrophage-Mannose Receptor (MMR) and
Mannose-Binding Lectin (MBL). MMR is a transmembrane protein, reported to bind to the terminal mannose, fucose and GlcNAc on the surface of pathogens, including bacteria, fungi,
76 | P a g e
and viruses. MMR binds to the influenza glycoproteins HA and NA, and is reported to mediate entry of influenza virus into macrophages [174, 176, 253, 254]. MBL, a soluble serum collectin, has been shown to directly bind H9N2 and H3N2 influenza strains [175].
The third type of interaction includes proteins that play a role in virus internalization or adhesion. Virus binding to receptor tyrosine kinases, such as epidermal growth factor receptor
(EGFR) has been shown to trigger intracellular signals, leading to formation of clathrin-coated pits, which promote influenza virus uptake [32]. Galectin (galactose-recognizing lectin) is reported to directly bind to the influenza virus surface glycoproteins. While this appears to decrease infectivity in vivo [255], in vitro studies suggest that binding of human and avian viruses to target cells is improved in the presence of galectin [256]. In this study, we compared the ability of the three different groups of host factors to capture influenza type A and B viruses.
3 Experimental procedures
3.1 Viral preparations and capture molecules
Influenza viruses were originally obtained from the Centers for Disease Control and
Prevention. The growth, characterization of HA titer, threshold cycle (Ct) values, and fluorescent-focus forming units have been described previously in Chapter 2 Table 2.1 and are summarized in Table 3.1. The biotinylated synthetic glycans (Figure 3.2) were obtained from the Consortium of Functional Glycomics (CFG). Purified proteins were obtained from the following sources: fetuin and asialofetuin (Sigma-Aldrich Co., USA); DC-SIGN, DC-SIGN receptor, mouse DC-SIGNR1/Fc, human MBL, and EGFR/Fc (Sino Biological Inc., China); human MMR/CD206 (R&D systems, USA); and galectin-1 (Novus Biologicals LLC., USA).
Fetuin was biotinylated with EZ-Link sulfo-NHS-LC-biotin (Pierce, USA) according to the manufacturer's instructions.
3.2 Capturing assay
Biotinylated synthetic glycans and biotinylated fetuin were dissolved in phosphate buffered saline (PBS) buffer (8.1 mM Na2HPO4, 1.5 mM KH2PO4, 128 mM NaCl, 2.7 mM KCl;
77 | P a g e
pH 7) and immobilized on streptavidin coated plates (SigmaScreen™, Sigma-Aldrich, USA) overnight at 4ºC. Host recombinant proteins at 2.5 μg/ml were immobilized overnight at 4ºC on hydrophilic microtiter plates (Microfluor® 2, Thermo Scientific, USA) using 100 mM bicarbonate/carbonate buffer at pH 9. After coating, all plates were washed with cold PBS and blocked with 2% BSA in PBS for 1 hour at 4ºC. After blocking, plates were washed and incubated with 1/30 dilutions of virus for 1 hour. Virus binding was evaluated at 4°C where the viral neuraminidase is not active [95], to avoid the need to add NA inhibitors. The plates were washed with cold PBS to remove unbound virus and virus binding was determined by measuring
NA activity at 37⁰C. NA activity reaction buffer consisted of PBS at pH 6.5 containing 0.1 mM
CaCl2 and 10 µM fluorogenic substrate 2′-(4-methylumbelliferyl)-α- D -N- acetylneuraminic acid
(MUNANA; Sigma-Aldrich, Inc.; USA). Reactions were incubated for 3 hours at 37ºC. Samples were adjusted to pH 9.5 using 0.7 M N-cyclohexyl-2-aminoethanesulfonic acid (CHES) buffer.
Fluorescence of the released 4-methylumbelliferone was measured with a FLX-800 fluorimeter
(BioTek, USA) using excitation and emission wavelengths of 350 and 460 nm, respectively. We used fetuin and MBL to determine the limit of detection of our capturing assay, using dilutions of virus: 1/30, 1/50, 1/90, 1/150 1/270, and 1/450.
Figure 3.2. Structures of the glycans used in this study. All the synthetic glycans are linked to biotin via the succinimidyl 6-(biotinamido) hexanoate (-LC-LC-Biotin) as described in Core resources at http://www.functionalglycomics.org. The typical glycosylation patterns found in fetuin [Takasaki, 1986] and typical high mannose core found in viral glycoproteins [Keil 1985, Roberts 1993]. Figure generated using Chemdraw.
78 | P a g e
4 Results
The receptor binding preference of six influenza A strains (H1N1 and H3N2) and three influenza B strains (Table 3.1) was examined, including closely related oseltamivir-sensitive influenza strains (indicated by –S) and oseltamivir-resistant influenza strains (indicated by –R).
All of the influenza preparations displayed similar levels of NA activity (ranging from about
40,000 to 53,000 RFU). Ct values, which measure viral genomes, were also fairly similar, ranging from 23 to 28. However, HA titers ranged from 40 to 320, and fluorescent focus forming units were even more variable, ranging from about 104 to 109 viable influenza particles per mL.
In summary, the virus samples tested here demonstrated equivalent NA activity and Ct values, suggesting similar levels of viral genetic material and NA protein were present in each sample.
However, biological activities, including host receptor recognition (HA titers) and infectivity (FFU values), varied greatly.
Table 3.1. Summary of characteristics of influenza strains and recombinant NA protein evaluated in this study Fluorescent NA act. Virus strain HA Titera focus forming Ctc RFUd units per mLb H1N1 strains A/California/07/09 H1N1pdm/09-S 320 7 x 108 23 53,000 A/North Carolina/39/09 H1N1pdm/09-R 80 5 x 106 25 49,000 A/Brisbane/59/07-Like H1N1/07-R 160 3 x 107 25 47,000 A/Florida/21/08 H1N1/08-R 320 2 x 106 25 40,000 H3N2 strains A/Brisbane/10/07-Like H3N2/07-S 40 3 x 106 23 53,000 A/Texas/12/07 H3N2/07-R 160 4 x 106 25 53,000 Type B strains B/Brisbane/60/08 Vict/06-S 320 5 x 107 28 47,000 B/Memphis/20/96 Vict/96-R 320 9 x 103 25 45,000 B/Florida/04/06 Yama/08-S 320 9 x 105 27 46,000 a : HA titers, assessing agglutination of red blood cells were determined using a standard WHO procedure b : Infectivity was determined using the fluorescent focus-forming assay. c: Ct values were determine by a standard WHO PCR assay using primers for amplification of matrix, nucleocapsid protein or non-structural 1 genes. We used 1,000-fold dilution of viral samples as detailed previously in Chapter 2 Table 2.1 d:NA activity for virus at 1 to 30 dilution, obtained as detailed in Methods.
79 | P a g e
4.1 Influenza binding to synthetic sialylated receptors
The ability of sialylated receptors to capture influenza virus was determined using streptavidin-coated plates. We tested biotinylated synthetic glycans displaying monoantennary terminal NeuAcα2-3Gal or NeuAcα2-6Gal (Figure 3.2). Additionally, we used biotinylated fetuin as a positive control. Fetuin is a highly sialylated protein, up to 9% total weight of fetuin can be attributed to NeuAc, displaying a glycosylation pattern with terminal multiantennary NeuAcα2-
3Gal or NeuAcα2-6Gal (Figure 3.2) [117]. Immobilized synthetic glycans and fetuin were incubated with whole viral particles on ice to inhibit NA activity. The plates were washed, and bound virus was detected by measuring NA activity expressed as relative fluorescent units
(RFU). As a negative control, we used virus-free samples. Since the RFU obtained in this study spanned in a wide range of values, a logarithmic scale was used in the y-axis in order to represent very low values, and percent (%) capture is indicated for values, which were significantly greater than the corresponding no virus control (Figure 3.3).
Fetuin, which displays both long oligosaccharides with terminal NeuAcα2-3Gal and
NeuAcα2-6Gal, was able to capture between 4 to 14% of H1N1 virus (Figure 3.3A-B), and 5 to
8% of type B virus (Figure 3.3D). However, while 4% of H3N2/07-S was captured, capture of the closely related oseltamivir-resistant strain H3N2/07-R was not significantly greater than the no virus control (Figure 3.3C). In general, capture with fetuin was better than all of the synthetic glycans. When synthetic receptors were compared, NeuAcα2-6Galβ1-3GlcNAc was overall the most effective capturing agent for influenza A, H1N1 viruses. Influenza A H3N2 and influenza B viruses, for the most part, bound equally to all synthetic glycans (Figure 3.3C-D), with the exception of Vict/96-R strain, where 5% of bound to NeuAcα2-3Galβ1-3GlcNAc, compared to
3% to the corresponding NeuAcα2-6 glycan (Figure 3.3D). Overall, the highly sialylated fetuin protein was the best capturing receptor for type A (H1N1 and H3N2) and type B (Victoria and
Yamagata) influenza strains; however, H3N2/07-R did not bind well to any of the sialylated receptors tested here.
80 | P a g e
Figure 3.3. Binding of influenza virus to synthetic sialylated glycans. Influenza strains are paired by similarity in different panels. A, pandemic influenza H1N1pdm/09 strains; B, influenza H1N1 (from 2007 and 2008); C, influenza H3N2 strains; D, influenza type B. Immobilized sialylated glycans were used to capture different influenza virus strains. Bound virus was detected by measuring viral NA activity. NA activity of total virus added to the wells is shown as Total NA act. . As a negative control, we used spent, virus-free media from mock infected Madin-Darby canine kidney (MDCK) cells which were used to propagate the viruses. Conformation of sialylated glycans used in this study are shown in the figure legend where sialic acid is represented as a gray diamond (♦), Galactose as an open circle (O), GlcNAc as a black square (■). The plots show at least 3 repeats of independent experiments and the error bar indicate standard deviation (SD). Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5. Percentages of captured viruses are shown for the receptors able to capture virus with statistical significant difference from negative control with P ≤ 0.01.
4.2 Influenza binding to protein host factors
We also tested the ability of host proteins immobilized on hydrophilic microtiter plates to capture influenza virus. We used fetuin, which displays terminal NeuAcα2-3 and NeuAcα2-6 glycans, as a positive control and compared binding to a non-sialylated protein, asialofetuin. For all viruses, more binding was seen when fetuin was immobilized on the hydrophobic plates compared to biotinylated fetuin immobilized on streptavidin plates. For example 14% of
H1N1pdm/09-S was captured with biotinylated fetuin (Figure 3.3A), while 23% was captured when fetuin was immobilized on the hydrophobic plates (Figure 3.4A). The increased capture
81 | P a g e
with fetuin could be due to the receptor alterations resulting from the biotinylation process, or a greater binding capacity or receptor display on the hydrophobic plates. All viruses bound fetuin better than asialofetuin, except for the H3N2/07-R strain which showed little capture by either molecule (Figure 3.4C). H3N2/07-R also bound poorly to the synthetic NeuAc-containing glycans (Figure 3.3C), suggesting capture may not be mediated by NeuAc.
Additionally, we considered protein factors, such as DC-SIGN, DC-SIGNR, mouse DC-
SIGN, MMR and MBL, which bind to high mannose glycosylation patterns characteristic of viral glycoproteins. For the DC-SIGN analogs, mouse DC-SIGN (mDC-SIGN) was the best capturing molecule binding 5 to 17% of H1N1 viruses, 4 to 11% of H3N2 strains and 4 to 7% of type B influenza viruses (Figure 3.4). Binding to MBL and MMR was similar for all influenza strains, where MBL was a slightly better capturing factor than MMR. MBL bound 7 to 20% of H1N1 viruses, 12 to 17% of H3N2 strains and 8% of type B Yamagata and Victoria viruses (Figure
3.4). MBL receptor was the best capturing factor among all mannose binding receptors in these assays. We observed that type A oseltamivir-sensitive strains bound better to mannose-binding factors than the closely related oseltamivir-resistant strains (white columns versus black columns in Figure 3.4A, 3.4C and 3.4D).
Finally, we considered protein factors that are reported to play a role in the interaction between influenza and host cells. EGFR is reported to be involved in viral internalization [32], while galectin is reported to block viral adhesion to host cells [255, 256]. We observed that
EGFR can capture 5 to 15% of H1N1 viruses, 2 to 4% of H3N2 viruses and 7% of influenza B
Yamagata and Victoria viruses (Figure 3.4). Binding to galectin was similar to EGFR for type A influenza strains, where maximum binding was observed to be up to 12% of H1N1/09-S.
Overall, the results of our binding studies show that fetuin and MBL are the best capturing factors tested in this study.
82 | P a g e
Figure 3.4. Capture of influenza virus by recombinant host proteins. Influenza strains are paired by similarity and capture was assessed by detecting NA activity as described in Figure 3.3. The plots show at least 3 repeats of independent experiments and the error bar indicate SD. Percentages of captured viruses are shown for the receptors able to capture virus with statistical significance compared to negative control at P ≤ 0.02.
4.3 Influenza limit of detection using Fetuin and MBL with in a capturing assay
We performed virus dilution studies to determine the viral limit of detection using the two best capturing factors, fetuin and MBL. Figure 3.5 shows the results using fetuin as a capturing molecule. Pandemic influenza A H1N1 strains (from 2009) were detected with fetuin capture at dilutions up to 270 to 450-fold (Figure 3.5A). H1N1 (from 2007 and 2008) and type B influenza strains were detected with fetuin capture up to a 150-fold dilution. Influenza H3N2/07-S strain was detected up to a 270-fold dilution. Binding of H3N2/07-R strain to fetuin was only detected at a 30-fold dilution.
83 | P a g e
Figure 3.5. Evaluation of the limit of Figure 3.6. Evaluation of the limit of detection using fetuin as a capturing detection using MBL as a capturing molecule for influenza virus. Influenza molecule for influenza virus. Influenza strains are paired by similarity in different strains are paired by similarity in different panels and capture was assessed by panels and capture was assessed by detecting NA activity as described in Figure detecting NA activity as described in Figure 3. The NA activity of virus at 1/30 dilution 5. The plots show at least 3 repeats of before capturing step is shown as Total NA independent experiments and the error bar activity. The plots show at least 3 repeats of indicate SD. The maximum dilution of virus independent experiments and the error bars giving statistical significant difference from indicate SD. Statistical differences were negative control is indicated with * for a P ≤ calculated by the two-tailed Student's t-test 0.02 and ** for a P ≤ 0.008. using GraphPad Prism™ 5. The maximum dilution of virus giving statistical significant difference from negative control (P ≤ 0.02) is indicated with *.
Capture with MBL was better or equal to fetuin (Figure 3.6), where the pandemic H1N1 strains were detected to a 450-fold dilution (Figure 3.6A), H1N1 (from 2007 and 2008) were detected to a 270 or 450-fold dilution (Figure 3.6B), H3N2 influenza strains were detected to a
450-fold dilution (Figure 3.6C), and influenza type B strains both Victoria and Yamagata were detected at 270-fold dilution (Figure 3.6D). Most significant was the ability of MBL to detect
H3N2/07-R down to 450-fold dilution, compared to a 30-fold dilution limit using fetuin.
84 | P a g e
5 Discussion
We used invariable characteristics of influenza biology (HA and NA) to capture and detect influenza virus. The ability of influenza HA to bind to sialylated receptors has been broadly studied. While it was thought that human H1 only recognized NeuAc attached in via α2-6 linkage and avian H1 only recognized the α2-3 linkage, recent data suggest otherwise [99, 100]. Both type A and type B human influenza virus are capable of binding terminal NeuAcα2-3 oligosaccharides and some strains prefer binding to NeuAcα2-3 over NeuAcα2-6 terminal sugars [99, 100, 102, 257]. Interestingly, while recombinant human HA proteins (H1pdm, seasonal H1 and H3) display exclusive affinity to α2-6 sugars, the corresponding viral particles bound to α2-3 sugars with comparable affinity to α2-6 sugars [102-104]. We also observed statistically significant binding of H1 and H3 viruses to α2-3 glycans (Figure 3.3A-C), and largely, the different viral isolates did not display a marked preference of any of the synthetic glycans. However, little to no binding of H3N2/07-R strain to synthetic receptors (Figure 3.3C) or to highly sialylated fetuin protein with multi-antennary glycan chains (Figure 3.4) was observed. This was not due to lack of HA activity, since the HA titer of H3N2/07-R was 4-times greater than the HA titer of H3N2/07-S (Table 3.1). Sequence alignment of the HA protein of
H3N2/07-S, which binds strongly, with H3N2/07-R and binds weakly to sialylated receptors, shows only two differences: Ala138Ser and Pro194Leu. Previously, weak affinity binding of HA to NeuAc was related to the accumulation of N-glycosylation sites on the globular head of the
HA near substrate binding site [258, 259]. The numerous glycosylations at the HA head could shield the substrate binding site, changing the specificity and recognition of NeuAc [260]. Both residues Ala138Ser and Pro194Leu are located at the globular head of HA near NeuAc binding site. It is possible that H3N2/07-R strain with those mutations displays an additional glycosylation site, which could block NeuAc substrate binding. No additional N-glycosylation sites on H3N2/07-R HA were found using NetNGlyc 1.0 server
(http://www.cbs.dtu.dk/services/NetNGlyc/), which predicts N-glycosylation sites, and only one
85 | P a g e
additional O-linked glycosylation was predicted on H3N2/07-R by NetOGlyc 3.1 Server
(http://www.cbs.dtu.dk/services/NetOGlyc/). Alternatively, the differences in binding affinity of closely related strains H3N2/07-S and H3N2/07-R could be due to changes in the NeuAc binding site. Even though our results show poor binding of H3N2/07-R to pure glycans in a biochemical assay, these strains grow well in cell culture having a high HA titer (160) and high
FFU value (4 x 106; Table 3.1).
In our studies using mammalian-expressed recombinant proteins as capturing agents
(Figure 3.4), we could not distinguish the ability of the mammalian proteins to capture the virus through their lectin-like activity from the ability of the whole virus to bind to NeuAc residues displayed on the protein. Studies with fetuin and asialofetuin show that binding of influenza strains to asialofetuin is significantly reduced compared to sialylated fetuin, suggesting that the loss of NeuAc reduces oligosaccharide binding. However, asialofetuin was still able to capture 1 to 3% of the virus (Figure 3.4). The influenza HA binding site is able to recognize additional residues in the oligosaccharide chain [10], and it is possible that the weak binding to asialofetuin is due to the HA recognition of Gal or Glc residues present in asialofetuin. The EGFR interaction with influenza is suggested to be mediated by NeuAc, and not viral recognition of the protein itself [32]. EGFR sequence analysis of potential glycosylation sites show up to 12 potential sites located on the extracellular domain of the protein [261]. Analysis of the binding patterns of
H3N2/07-R, which did not bind efficiently to NeuAc, provides insight into virus-mediated binding versus host protein-mediated binding. H3N2/07-R binds fetuin as well as asialofetuin and EFGR.
This suggests that binding to EGFR may be mediated by NeuAc, and not by an intrinsic affinity of the virus for protein domains on EGFR. In contrast, more binding to H3N2/07-R is seen for
MMR, MBL and some DC-SIGN family members, suggesting that this binding is primarily due to the ability of the host proteins to recognize the virus.
The role of galectin in influenza infection is not well understood. In vitro, human galectin-1 promotes viral internalization, and blocks adhesion to host cells [256]; however, it enhances
86 | P a g e
survival of mice in vivo [255]. In addition, galectin has been reported to bind to influenza surface glycoproteins and inhibit HA activity [255, 262]. Like EGFR, galactin-1 was not efficient at capture of H3N2/07-R. Furthermore, the other influenza strains bound galectin about as well as
EGFR (Figure 3.4). From these studies we cannot rule out that the observed binding is mediated by viral recognition of NeuAc, instead of intrinsic affinity for galectin for the virus.
High mannose cores (Figure 3.2) are recognized as a pathogen-associated molecular pattern (PAMP) and plays a significant role in innate immune response to viral infection [157], and can be found attached to several residues on influenza HA [155, 156]. Mouse DC-SIGN, which recognizes dense arrays of mannose glycans better than individual mannose residues
[250-252], was able to capture virus slightly better than human DC-SIGN analogs (Figure 3.4).
Interestingly, the H3N2/07-R strain, which was not captured well using sialylated receptors, bound well to all DC-SIGN analogs (Figure 3.4C). MBL and MMR, which strongly bind to terminal mannose residues [253, 254], were able to capture more virus than DC-SIGN analogs, which do not specifically recognize terminal mannose.
Overall, fetuin and MBL were the best agents for viral capture. The virus binds to fetuin, while MBL binds to the virus. Fetuin-capture virus at 150-450 fold dilutions (Figure 3.5), while
MBL-capture virus at 270-450 fold dilutions (Figure 3.6). This approach could be used to develop quick and reliable diagnostic tests for influenza. Such tests could significantly reduce the disease burden by rapidly and accurately identifying infected individuals and helping control the transmission of influenza virus. The use of intrinsic viral characteristics provides a great advantage over antibody-recognition, since binding to host factors is not subject to antigenic shift or drift.
The results of this chapter can be found in :
Gallegos KM, and Weiss AA (2013) Capturing influenza virus with host receptors for detection.
Manuscript submitted for publication.
87 | P a g e
CHAPTER 4: SHIGA TOXIN BINDING
1. Abstract
Immunologically distinct forms of Shiga toxin (Stx1 and Stx2) display different potencies and disease outcomes, likely due to differences in host cell binding. The glycolipid globotriaosylceramide (Gb3) has been reported to be the receptor for both toxins. While there is considerable data to suggest that Gb3 can bind Stx1, binding of Stx2 to Gb3 is variable.
We used isothermal titration calorimetry (ITC) and enzyme-linked immunosorbent assay
(ELISA) to examine binding of Stx1 and Stx2 to various glycans, glycosphingolipids and glycosphingolipid mixtures in the presence or absence of membrane components, phosphatidylcholine and cholesterol. We have also assessed the ability of glycolipids mixtures to neutralize Stx-mediated inhibition of protein synthesis in Vero kidney cells.
By ITC, Stx1 bound both Pk (the trisaccharide on Gb3) and P (the tetrasaccharide on globotetraosylceramide, Gb4), while Stx2 did not bind to either glycan. Binding to neutral glycolipids individually and in combination was assessed by ELISA. Stx1 bound to glycolipids
Gb3 and Gb4, and Gb3 mixed with other neural glycolipids, while Stx2 only bound to Gb3 mixtures. In the presence of phosphatidylcholine and cholesterol, both Stx1 and Stx2 bound well to Gb3 or Gb4 alone, or mixed with other neutral glycolipids. Pre-incubation with Gb3 in the presence of phosphatidylcholine and cholesterol neutralized Stx1, but not Stx2 toxicity to Vero cells.
Stx1 binds primarily to the glycan, but Stx2 binding is influenced by residues in the ceramide portion of Gb3 and the lipid environment. Nanomolar affinities were obtained for both toxins to immobilized glycolipids mixtures, while the effective dose for 50% inhibition (ED50) of protein synthesis was about 10-11 M. The failure of preincubation with Gb3 to protect cells from
88 | P a g e
Stx2 suggests that in addition to glycolipid expression, other cellular components contribute to toxin potency.
2. Introduction
Escherichia coli O157:H7 is the most common serotype of Shiga toxin-producing E. coli isolated from patients in the United States. It is estimated to cause 110,000 cases, mostly among children and the elderly, and 3,200 hospitalizations annually in the United States, costing approximately 400 million dollars [263, 264]. This pathogen causes food-borne disease with symptom severity that varies from mild diarrhea to hemorrhagic colitis, and potentially to life- threatening Hemolytic Uremic Syndrome (HUS) [265]. Shiga toxin (Stx), the most important virulence factor of E. coli O157:H7, is responsible for the life-threatening complications following infection. Stx is an AB5 toxin consisting of a single A subunit associated with a pentamer of identical B subunits. This pentamer binds to the glycosphingolipid globotriaosylceramide (Gb3) in host cell membranes [188, 266-268] and delivers the A subunit into the cytoplasm. In the cytoplasm, the enzymatically active A subunit inhibits protein synthesis by cleaving an adenine nucleotide from 28S RNA within the 60S ribosomal subunit, preventing tRNA binding and protein synthesis [269, 270].
There are two immunologically distinct forms of Stx: Stx1 and Stx2. They share 56.8% amino acid sequence identity [271, 272]. In epidemiological studies, Stx2 is more often associated with severe disease outcome and development of HUS than Stx1 [265]. In animal models, Stx2 is 100- to 400-fold more potent than Stx1 [273-275]. Differences in host cell receptor binding between Stx1 and Stx2 appear to mediate the differences in potency in vivo and in vitro [274, 276-279]. Shimizu et al. reported that a chimeric toxin with the Stx2A subunit associated with the Stx1B-pentamer was 2-fold more toxic to mice than wild type Stx1 and 50- fold less potent than wild type Stx2, suggesting that the A subunit does not significantly contribute to potency in vivo, while the B-pentamer play a more significant role [267]. These data suggest that Stx potency might be due to a differential targeting or affinity in binding to host cell
89 | P a g e
receptors. When Stx1 or Stx2 is administered to mice, Stx1 stays predominantly the lungs without causing pathology while Stx2 mainly targets the kidneys [274, 280]. It has been suggested that Stx1 might bind to Gb3 variants in the lungs, preventing it from reaching more susceptible organs such as the kidneys, whereas Stx2 binds preferentially to Gb3 variants in kidney tissue.
Stx binding to the Pk trisaccharide (Galα1-4Galβ1-4Glc) present in Gb3 occurs primarily through hydrogen bonds between the hydroxyl groups on the sugars. High affinity is achieved through avidity, by engaging multiple binding sites on the toxin. The Stx1 B-pentamer has 3 Pk trisaccharide binding sites per subunit, or 15 sites total per holotoxin [281]. In contrast, the binding sites for Stx2 are less well defined, but the binding interactions have been modeled
[282, 283]. Interestingly, binding studies using receptor mimics show that Stx1 binds with higher affinity to the Pk trisaccharide than Stx2 [276, 284-287]. Published data demonstrate different and selective binding preferences of Stx1 and Stx2 to synthetic glycans. Stx1 shows a preference for binding native Pk while Stx2 binds better to an N-acetylated analogue of Pk (NAc-
Pk) [277, 285, 288]. Native Pk trisaccharide is found on glycolipid Gb3, while NAc-Pk is found on proteins, but no glycolipids with NAc-Pk are known to exist in nature.
Native Gb3 is found on the lipid rafts (detergent-insoluble glycolipid-enriched domains) in host cell membranes. Lipids rafts are composed of (glyco)sphingolipids, glycerophospholipids, and cholesterol. Stx2 variants, such as porcine edema disease toxin (Stx2e), have been reported to bind to glycosphingolipid globotetraosylceramide (Gb4), which contains an additional residue, GalNAc, attached to the Pk of Gb3 [289, 290]. In a recent report, Gb3 was found to be present in low quantities in colonic epithelial cells in vivo; whereas Gb4 was found abundantly
[291]. Low affinity binding of Stx1 to Gb4 has been reported [291], suggesting that Stx1 could bind to these glycolipids in host cells membranes. However, the true functional receptor of Stx remains unknown. It is not clear if Gb3 is the main factor mediating Stx binding to host cells, and in vitro binding affinities do not correlate with cellular or in vivo toxicity. Previous data shows that
90 | P a g e
Stx affinity for Gb3 is in the nanomolar range while cellular and in vivo toxicity are in the picomolar range, suggesting other factor might also play a role of Stx toxicity in vivo and at cellular level [273, 276, 284].
Recently, it has been reported that lectin binding was enhanced in the presence of glycolipid mixtures as compared to the binding to single glycolipids [292]. Considering that glycolipids are naturally found in the cell membrane in mixtures and in combination with phospholipids and cholesterol [293, 294], Stx binding in vivo might involve more than one glycolipid, and the presence of cholesterol and phospholipids.
The objective of this study was to gain insight into the receptor preferences for Stx1 and
Stx2. We examined binding of Stx1 and Stx2 to various glycans, glycolipids and glycolipid mixtures by ITC or ELISA in the presence or absence of phosphatidylcholine (PC) and cholesterol (Ch). The findings of this study have clarified the differences in binding of Stx1 and
Stx2.
3. Materials and Methods
3.1. Production of recombinant Stx toxoids and B-pentamers
Toxin-encoding genes were PCR amplified and cloned into the expression plasmids, as outlined in Table 4.1. The sequence of all inserts was verified. To generate the Stx2 toxoid expression construct (pTSG218), the inactivated stx2 operon from pNR100 [295] was cloned as single PCR product. To generate the Stx1 toxoid expression construct, pTSG214 containing the stx1A and B genes in tandem, tyrosine 77 and glutamic acid 167 of stx1A were sequentially replaced with serine and glutamine, respectively using the QuickChangeTM protocol
(Stratagene) generating pTSG213. stx1B was excised from pTSG211 with XbaI and NotI and cloned into the NotI and SpeI site of pTSG213.
91 | P a g e
Table 4.1. Plasmids and primers used in this study
Plasmid Genotype Vector / PCR Template [Reference] pTSG210 Stx1A-WT pETSecS3 [41] / pMFUC-17 [Flagler et al., 2010] pTSG211 Stx1B-WT pETSecS3 / pSW09 [Wen et al., 2006] pTSG212 Stx1A-Y77S pTSG210 [This study] pTSG213 Stx1A-Y77SE167Q pTSG212 [This study] pTSG214 Stx1A-Y77SE167Q + Stx1B-WT pTSG213 [This study] pTSG218 Stx2A-Y77SE167Q + Stx2B pETSecS3 / pNR100 [Wen et al., 2006] pTSG230 Stx2B-WT pETSecS3 [Millen et al., 2010] / pMFCU-21 [Flagler et al., 2010] Cloning Primers Name Sequence 5’ Stx1A NdeI AACATATGATGAAAATAATTATTTTTAGAGTGC 3’ Stx1A SpeI ATACTAGTTCAACTGCTAATAGTTCTGCGC 5’ Stx1B NdeI AACATATGATGAAAAAAACATTATTAATAGCTGC 3’ Stx1B SpeI ATACTAGTTCAACGAAAAATAACTTCGCTG 5’ Stx2A NdeI AACATATGATGAAGTGTATATTATTTAAATGGG 3’Stx2A SpeI ATACTAGTTCAGTCATTATTAAACTGCACTTC 5’Stx2B NdeI GGAATTCCATATGAAGAAGATGTTTATGGCGG 3’Stx2B SpeI GGACTAGTTCAGTCATTATTAAACTGCACTTCAG Site-Directed Mutagenesis Primers Name Sequence GGTTTAATAATCTACGGCTTATTGTTGAACGAAATAATTTAAGTGTGAC 5’ Stx1A-Y77S AGGATTTGTTAACAG CTGTTAACAAATCCTGTCACACTTAAATTATTTCGTTCAACAATAAGCC 3’ Stx1A-Y77S GTAGATTATTAAACC 5’ Stx1A-E189Q CGGTTTGTTACTGTGACAGCTCAGGCTTTACGTTTTCGGC 3’Stx1A-E189Q GCCGAAAACGTAAAGCCTGAGCTGTCACAGTAACAAACCG
Proteins were expressed from cold-induced cultures as previously described [277, 296], with the following modifications. Briefly, logarithmic phase cultures were cooled to 8°C; expression of recombinant toxoid and protein folding genes was induced by addition of IPTG
(0.1 mM) and ethanol (2%), respectively. After overnight incubation with shaking at 20°C, the cells were harvested by centrifugation, and lysed by gentle shaking with 4M urea for 30 minutes.
Cellular debris was removed by centrifugation. The extract was dialyzed, and concentrated as a
40-70% ammonium sulfate fraction. Toxoids were further purified using combinations of AffiGel
Blue affinity chromatography (Bio-Rad, CA), ion exchange, or size exclusion chromatography.
92 | P a g e
Pigeon egg white affinity chromatography [296] was used for Stx1 toxoid. Protein was quantified using bicinchoninic acid protein assay (Pierce, IL). Purity of toxoid was verified by the presence of only two bands corresponding to the A-and B-subunits on Coomassie stained 8-16% polyacrylamide gels (Lonza) loaded with 1g of protein.
3.2. Isothermal Titration Calorimetry (ITC)
ITC experiments were performed in a Microcal VP-ITC microcalorimeter at 25°C in buffer containing 20 mM HEPES at pH 7.4 and 150 mM NaCl. Stx1B and Stx2B were dialyzed into this buffer, and powdered Pk trisaccharide and P tetrasaccharide glycans were resuspended in dialysate to achieve a buffer match. All experiments were performed with Stx B-subunits in the microcalorimeter cell at 238-300 μM concentration, and glycans in the syringe at 50 mM concentration. The titrations consisted of a total of forty 7-μl injections, spaced 120 seconds apart. Protein concentrations were determined based on the UV absorbance at 280 nm and molar extinction coefficients of the Stx1B and Stx2B monomers (8,605 M-1cm-1 and 14,105 M-
1cm-1, respectively). Data were analyzed in ORIGIN using a one-site binding model with fixed n=1 per B subunit (the fixed parameter was required to achieve convergence of the fit). The Kd values reported are the average of two replicates.
3.3. Glycan array studies
Stx1 (2.84 μM) and Stx2 (0.64 μM) toxoids (obtained from the BIODEFENSE AND EMERGING
INFECTIOUS DISEASES RESEARCH RESOURCES REPOSITORY, Manassas, VA) were submitted to the Consortium for Functional Glycomics (CFG) to assess glycan binding specificity. The
Mammalian Printed Array Version 4.1 holds 465 different glycans consisting of natural and synthetic mammalian glycans. Toxin binding was detected using rabbit polyclonal antibody to
Stx (Meridian Bioscience, Cincinnati, OH) and fluorescently labeled anti-rabbit IgG Alexa 488 antibody which was supplied by the CFG. The array consists of six replicates of each glycan, and relative binding was expressed as mean relative fluorescence units (RFU) of four of the six
93 | P a g e
replicates after removal of the highest and lowest values. Binding data can be accessed at the
CFG website (http://www.functionalglycomics.org/).
3.4. Glycolipid ELISA
Glycolipids and lipids (Table 4.2) were purchased from Matreya Inc. (Pleasant Gap, PA).
Pure glycolipids were suspended in chloroform and diluted in methanol as previously described
[292]. Mixtures of glycolipids were prepared in a molar ratio of 1:1. Mixtures of glycolipids with cholesterol (Ch) and phosphatidylcholine (PC) were prepared in a molar ratio of 1:3:3 as previously described [276]. Single or mixed glycolipids with or without Ch and PC were added to wells of hydrophobic Microtiter® plates (Microfluor® 1, Thermo scientific) and dried for 30 hours in a fume hood. As negative controls, methanol alone, PC, Ch or PC+Ch were added to wells. In all experiments, background RFU values obtained in methanol were subtracted from each value.
Except were indicated, all steps were performed at 4°C. Prior to use, the plates were blocked for
1 hour with phosphate buffered saline (PBS; 8.1 mM Na2HPO4, 1.5 mM KH2PO4, 128 mM NaCl,
2.7 mM KCl), pH 7.4, containing 2% (w/v) bovine serum albumin (BSA). Dilutions of Stx toxoid were added and incubated for 1 hour, followed by sequential incubation with primary antibody against Stx1 or Stx2 (rabbit polyclonal serum, Meridian Bioscience, Cincinnati, OH) and peroxidase-conjugated goat anti-rabbit IgG (MP Biomedicals, Solon, OH). Wash steps were carried out using cold PBS pH 7.4 containing 1% (w/v) BSA. Finally, plates were developed with
QuantaBlu® fluorogenic peroxidase substrate (Pierce, Rockford, IL) and read. Binding curves and analysis were performed using Prism 5.0 (GraphPad Software, La Jolla, CA).
3.5. Vero protection studies
Microtiter plates were coated with glycolipid as described above. Wells treated with methanol alone, PC + Ch alone, or not pre-treated served as negative controls. Unbound surfaces on the wells were blocked with Minimal Essential Medium 1X (Invitrogen™) supplemented with 10% fetal bovine serum, vitamins (Sigma-Aldrich™) and glutamine
(Sigma™), and washed with PBS. Stx1 and Stx2 (Biodefense and Emerging Infectious Diseases
94 | P a g e
Research Resources Repository, Manassas, VA) were serially diluted in PBS and added to the wells, starting with 10-8 M of toxin. The toxin was incubated at 37C for 1 hour. After incubation, the toxin was removed from the wells and added to tissue culture treated 96 well plates (Corning
Inc.™). The amount of residual toxin was determined as previously described [273, 297] by measuring protein synthesis inhibition using Luc2P Vero cells engineered to express destabilized luciferase [297] . Briefly, Luc2P Vero cells were added at 104 cells per well. After 4 hours of incubation at 37C and 5% CO2, the cells were washed with PBS and 25 l/well of
SuperLight luciferase substrate was added and luminescence was measured. The results were reported as percentage of maximum signal from PBS control cells incubated without any toxin.
The effective dose to inhibit 50% of protein synthesis (ED50) was calculated using the two points above and below the midpoint and normalized against the untreated control.
4. Results
4.1. Characterization of individual glycan binding sites by ITC
While the glycolipid Gb3 is commonly reported to be the receptor for both Stx1 and Stx2, the two toxins appear to have different receptor preferences. We used ITC to examine binding of
Stx1 and Stx2 to the Pk-trisaccharide expressed on Gb3. To avoid complications due to A- subunit interactions, binding studies were performed with purified B-pentamer. Stx1B bound to
Gb3 with a Kd of about 4 mM (Figure 4.1A), which is in good agreement with previously published studies using ITC [298] and mass spectrometry [299]. In contrast, no binding of Stx2B to Pk was detected under the experimental conditions tested (Figure 4.1B).
In previous reports, Stx1 and Stx2 binding to Gb4 was observed [284, 290, 300-302] and recently Stx1 has been reported to bind to Gb4 [291]. However, nothing is known about the number or affinity of single sites for Gb4. Stx1B bound to the P tetrasaccharide expressed on
Gb4 with a Kd of 12 mM (Figure 4.1C), with about 3-fold lower affinity compared to Gb3. These results demonstrate that Stx1 might recognize Gb4 as receptor. Like Pk, no binding of Stx2B to
95 | P a g e
P-tetrasaccharide was observed (Figure 4.1D), which suggested that the Kd of Stx2B for both glycans is much greater than 12 mM.
Figure 4.1. Binding of Stx1 and Stx2 to purified Pk trisaccharide and P tetrasaccharide by ITC. Glycans (50 mM) were titrated into a microcalorimeter cell containing 238-300 μM of Stx B-subunits. Stx binding interaction with Pk (A, B) and P tetrasaccharide (C, D). Both Stx1 B-subunits (A-C) and Stx2 B- subunits (B-D) raw heat signals (top) and integrated data from titrations (bottom) are shown.
4.2. Stx binding to synthetic glycans
To identify other possible glycan receptors, Stx1 and Stx2 toxoids were assayed by ELISA for binding to 465 different glycans by the Consortium for Functional Glycomics. To avoid exposure to the high concentrations of toxin typically used in binding studies, these studies were performed with genetically inactivated toxin. The two amino acid changes (Tyr77Ser and
Glu167Gln) abolish the enzymatic activity of the A-subunit, but do not affect binding mediated by the B-pentamer [295, 303].
No significant binding was detected for Stx2 at 0.64 μM (data not shown). Binding to Stx1 was detected. The top three hits for Stx1 (Figure 4.2, glycans 331, 402, and 120) resembled Pk trisaccharide, with Galα1-4Gal as the terminal sugars; however they differed from Pk at the third
96 | P a g e
sugar, which was GlcNAc instead of Glc (Figure 4.2). Interestingly, Stx1 did not display significant binding to glycan 121, containing the native Pk antigen. However, consistent with the above results, 7-fold more binding of Stx1 was observed to glycan 119 (the 6th best hit) which only differed from glycan 121 by the presence of N-acetylation at the third residue, suggesting
GlcNAc may be the preferred residue. However, in nature, this glycan (Galα1-4Galβ1-4GlcNAc) occurs in mammalian glycosylated proteins, but not on glycolipids.
Figure 4.2. Glycan array results for Stx1. Binding of Stx1 (2.84 μM) toxoid to the CONSORTIUM FOR FUNCTIONAL GLYCOMICS MAMMALIAN Array Version 4.1 with 465 different natural and synthetic mammalian glycans was assessed by ELISA. Displayed are the top three hits for Stx1 (glycans 331, 402, and 120). For comparison, also displayed is native Pk (glycan 121), glycan 119 which is attached using the same linker as native Pk, and glycan 120, which is attached with a different linker from 119. The symbolic representation of the compounds follows the CFG standards: galactose (Gal, white circle),glucose, (Glc, black circle), N-acetyl-glucosamine (GlcNAc, black square), mannose (Man, gray circle). X corresponds to β1-4GlcNΑcβ1-4GlcNΑcβ- LVΑNKT. Spacers used to couple the glycans to the array surface matrix: Sp0, -
CH2CH2NH2; Sp8, -CH2CH2CH2NH2; LVΑNKT, peptide (Leucine, L; valine, V; alanine, A; asparagine, N; lysine, K; threonine, T). Relative fluorescence units (RFU) signal is the mean of four independent experiments and error bars indicate Standard Deviation (SD).
Additionally, the linker used to attach the glycans to the array surface matrix can influence toxin binding [285, 304, 305]. Glycan 120 and glycan 119 share the identical glycan trisaccharide, but are attached with different linkers. A change from the Sp0 linker (-
CH2CH2NH2) to the Sp8 linker (-CH2CH2CH2NH2) increased Stx1 binding by 2-fold.
4.3. Stx binding to native Gb3 glycolipid
The failure of the glycan array to reveal binding of Stx1 to native Pk, and the inability to detect any ligands for Stx2 led us to examine binding to native glycolipids. In initial experiments, binding at several concentrations of Stx was assessed using pure Gb3 immobilized on
97 | P a g e
hydrophobic microtiter ELISA plates with incubations at 4C. The apparent dissociation constant
(Kd) of Stx1 binding to Gb3 was determined to be 4.2 nM (Figure 4.3) which is 10-fold lower than the 46 nM value reported with radio-labeled Stx1 and 48-fold lower than the 222 nM value reported with Surface Plasmon Resonance (SPR) [276, 284]. There are several explanations for the different apparent Kd values obtained in different studies. ELISA has been shown to be more sensitive than SPR [277, 306], possibly because the longer incubation periods in the static
ELISA allows the toxin to achieve optimal interacting conformation compared to the dynamic flow conditions of SPR [305]. Additionally, we incubated the plates at 4°C, while the SPR studies were done at room temperature.
Figure 4.3. Stx binding to pure Gb3. Stx1 (black squares, ■) and Stx2 (black circles, ●) toxoid binding affinity to Gb3 alone was assessed by ELISA at 4⁰C. Stx1 binding as fitted to a one-site specific binding model with Hill coefficients. Symbols represent experimental data, while lines represent the fitted model for that data analyzed with Prism5 (GraphPad software, La Jolla, CA). Values for Stx2 were not determined due to poor binding. The RFU signal is the mean of three independent experiments and error bars indicate SD.
In contrast a Kd for Stx2 binding to Gb3 was not determined due to the high concentration of toxin (above 1 μM) needed to reach saturated binding under these conditions (Figure 4.3).
Previous studies reported low affinity binding of Stx2 to Gb3 using radio-labeled Stx2 (Kd= 370 nM) and SPR (Kd= 1040 nM) [276, 284].
4.4. Stx binding to glycolipid complexes
Rinaldi et al. (2009) suggested that mixed glycolipid complexes may support better binding than pure glycolipids. Neutral glycolipids of the glucosylceramide family are synthesized by sequential addition of sugars to the ceramide core, culminating with the tetrasaccharide form,
Gb4 (Table 4.2). The glucosylceramides display a broad cellular distribution. In contrast, the glycolipid galactosyl ceramide (Gal-Cer), which is synthesized by a different pathway, is found
98 | P a g e
primarily on neuronal tissue [307, 308]. Since selective binding to Stx2 to NAc-Pk is reported, we also evaluated binding of Stx to asialo GM1 (aGM1) and asialo GM2 (aGM2) gangliosides
[277, 285]. The glycan portion of aGM1 (GalNAcβ1-4Galβ1-4Glc) is similar to NAc-Pk except for the β1-4 instead of α1-4 linkage of the terminal GalNAc residue; aGM2 is a derivative of aGM1 with an additional Gal residue added with a β1-3 linkage.
Table 4.2. Glycolipids used in Chapter 4
Abbreviation, Empirical Name Structure Product number Formula
Glucosyl ceramide Glc-Cer, 1521 Glc-Ceramide C48H93NO8 Lactosyl ceramide Lac-Cer, 1507 Gal1-4Glc-Ceramide C53H101NO13 Globotriaosyl ceramide Gal1-4Gal1-4Glc- Gb3, 1067 C H NO Ceramide trihexoside Ceramide 60 113 18 GalNAc1-3Gal1-4Gal1- Globotetraosyl ceramide Gb4, 1068 C H N O 4Glc-Ceramide 68 126 2 23 Gb3 with non-hydroxy fatty Gal1-4Gal1-4Glc- Gb3 –OH, 1513 C H NO acid side chain Ceramide 54 101 18 Gb3 with hydroxy fatty acid Gal1-4Gal1-4Glc- Gb3 +OH, 1514 C H NO side chain Ceramide 54 101 19 Lyso-globotriaosyl Gal1-4Gal1-4Glc- Lyso-Gb3, 1520 C H NO sphingosine Ceramide 36 67 17 Galactosyl ceramide Gal-Cer, 1050 Gal-Ceramide C48H93N08 GalNAc1-4Gal1-4Glc- Asialo GM gangliosides aGM , 1512 C H N 0 2 2 Ceramide 56 104 2 18 Gal1-3GalNAc1-4Gal1- Asialo GM gangliosides aGM , 1064 C H N 0 1 1 4Glc-Ceramide 62 114 2 23 We examined binding of Stx1 and Stx2 to the neutral glycolipids, alone or in combination
(Figure 4.4A). Stx1 (10 nM) bound to Gb3 and Gb4, but not to Glc-Cer, Lac-Cer, Gal-Cer, aGM1 or aGM2 (Figure 4.4A, white bars). Stx1 also bound to 1:1 mixtures of Gb3 and the other glycolipids, and some mixtures of Gb4. In contrast, strong binding of Stx2 was only observed for
Gb3 mixed with Glc-Cer, Lac-Cer or Gal-Cer (Figure 4.4A, black bars).
In mammalian cells, glycolipids in lipid rafts are arrayed in fluid membranes containing cholesterol (Ch) and phosphatidylcholine (PC). We also examined Stx binding to glycolipid mixtures in the presence of these other membrane components (Figure 4.4B). Individual glycolipids Glc-Cer, Lac-Cer, Gal-Cer, aGM1 and aGM2 failed to support binding of either Stx1 or
Stx2 even in the presence of Ch and PC (Figure 4.4B). However, the presence of Ch and PC
99 | P a g e
resulted in increased binding of both Stx1 and Stx2 to Gb3 and Gb4, and both toxins bound to a broader array of glycan mixtures. These initial studies were performed at 4C. Since membrane fluidity is much greater at physiological temperatures, we repeated these binding studies at
37C. Incubation at 37C only resulted in significantly increased binding to Gb3 and Gb4 reflected in higher RFU values (data not shown).
Figure 4.4. Binding of Stx1 and Stx2 to purified glycolipids and mixtures. Stx binding was assessed by ELISA at 10 nM for both Stx1 (white columns) and Stx2 (black columns) at 4⁰C. The RFU signal is the mean of three independent experiments and error bars indicate SD. Since different antibodies were used to detect Stx1 and Stx2, two axes are shown. (A) Binding of Stx1 and Stx2 to purified glycolipids and mixtures in absence of Ch and PC. Mixtures of glycolipids were prepared in methanol at a ratio of 1:1 and added at 200 ng of total glycolipid per well. (B) Binding of Stx1 and Stx2 to purified glycolipids and mixtures in the presence of Ch and PC. Mixtures were prepared in methanol at a ratio of glycolipid 1, glycolipid 2, cholesterol, phosphatidylcholine 1:1:3:3 and added at 200 ng of total glycolipid per well.
The apparent Kd of Stx1 and Stx2 for Gb3 and Gb4 in the presence of Ch + PC was assessed at 37C (Figure 4.5). The apparent Kd of Stx1 for Gb3 in the presence of Ch + PC was 6.4 nM, which is very similar 4.2 nM, the apparent Kd of Stx1 for Gb3 at 4C without Ch +
PC (Figure 4.3). However, the shape of the binding curves was very different, a reflection of the very different hill coefficients (nH), 1.4 for pure Gb3 at 4C (Figure 4.3) versus nH = 0.38 for binding to Gb3 with Ch and PC at 37C (Figure 4.5A). Since Stx1 has multiple binding sites for
Pk, the Hill coefficient of less than 1 seen for Gb3 in the presence of PC and Ch suggests different classes of apparent affinity of the Stx1 toxoid for binding to the plate. This reflects differing levels of avidity rather than differences in individual sites on the Stx1 B-pentamer, likely due to microheterogeneity in the lipid makeup at the plate surface. The avidity of Stx1 for Gb4
100 | P a g e
was nearly identical to Gb3 alone, with an apparent Kd of 3.9 nM compared to 6.4 nM (Figure
4.5A), but Gb4 supported less binding than Gb3. Binding of Stx1 to Gb3/Gb4 mixture displayed an apparent Kd (6.2 nM) very similar to that obtained with either glycolipid alone. Interestingly, in the presence of Ch and PC, Stx2 binding to Gb3, Gb4 or mixtures was very similar to Stx1, both in global affinity (Kd of 6.4 nM, 14 nM, and 3.2 nM, respectively) and displaying Hill coefficients of less than 1.
Figure 4.5. Stx binding to Gb3, Gb4 and Gb3/Gb4 mixture in the presence of cholesterol and phosphatidyl-choline. Stx1 (A) and Stx2 (B) toxoid binding was assessed by ELISA at 37⁰C. As negative controls, toxin was incubated in methanol, PC, Ch, or PC+Ch coated wells. In all experiments, background RFU values obtained in methanol were subtracted from each value. Binding curves were fitted to a one-site specific binding model with Hill coefficients. Symbols represent experimental data, while lines represent the fitted model for that data analyzed with Prism5 (GraphPad software, La Jolla, CA). The RFU signal is the mean of three independent experiments and error bars indicate SD.
The differences in Bmax for Gb4 compared to Gb3 or Gb3/Gb4 for both toxins is significant.
This suggests that the number of individual sites on both Stx B-pentamers that can bind Gb4 are presumably lower than the number of sites able to bind Gb3; resulting in a less stringent binding of the B-pentamer to Gb3 compared to Gb4. In support of this hypothesis, similar results
(glycans with identical apparent Kd but very different Bmax values) were observed for pertussis
101 | P a g e
toxin, an AB5 toxin with non-identical B-subunits known to possess structurally and functionally heterogeneous glycan binding sites [304].
4.5. Contribution of the ceramide to Stx binding
Figure 4.6. Stx binding to Gb3 analogs. Stx binding was assessed by ELISA at 10 nM for both Stx1 (A) and Stx2 (B) at 37⁰C. Gb3 -OH FA, with non- hydroxy Fatty Acid chain; +OH FA with hydroxy Fatty Acid chain. If not specified Gb3 is a standardized mixture that contains both variants with hydroxyl and nonhydroxyl fatty acid chains (Matreya Inc.). As negative controls, toxin was incubated in methanol, PC, Ch, or PC+Ch coated wells. In all experiments, RFU values obtained in methanol were subtracted from each value in order to define a base level. The RFU signal is the mean of three independent experiments and error bars indicate SD.
To determine if the sphingosine residues in the ceramide portion of Gb3 molecule played a role in binding to Stx2, we assessed binding to variants of Gb3 with or without the α- hydroxylated fatty acid (OH FA) in the ceramide (Figure 4.6). Binding of these variants was compared to the preparation that contains both variants, hydroxyl and nonhydroxyl fatty acid chains, used in Figures 4.3-4.5. Stx1 displayed similar binding to Gb3 regardless of presence of the ceramide hydroxyl or the presence of Ch and PC (Figure 4.6A). In the absence of PC and
Ch, Stx2 failed to bind Gb3 regardless of which form of ceramide hydroxyl was present, and in the presence of PC and Ch bound equally to Gb3 expressing either form of ceramide (Figure
4.6B). These results demonstrate that the hydroxyl residue in the fatty chain of the sphingosine part of Gb3 does not play a significant role in binding to either Stx1 or Stx2, and agree with previous reports that the OH FA variants of Gb3 display similar binding affinities for both Stx1 and Stx2 [300].
102 | P a g e
To investigate further the role of the ceramide in Stx binding, we evaluated binding to deacylated Gb3 (Lyso-Gb3). Lyso-Gb3 lacks a carbony group and one fatty acid chain (acyl group) in the sphingosine of Gb3 (Figure 4.7). Stx1 displayed about a third as much binding to
Lyso-Gb3 in the presence of Ch and PC (Figure 4.7A). Stx2 did not bind to Lyso-Gb3 in the presence or absence of Ch and PC (Figure 4.7B). These results demonstrate that either the presence of the ketone group or the acyl group in Gb3 is essential for binding to Stx2 at low concentrations of toxin. ELISA probing coated wells with an anti-Gb3 antibody suggest that about 2.8 times more Gb3 than lyso-Gb3 binds to the hydrophobic well in the presence of
+PC+Ch; therefore, reduced binding of Stx1 for lyso-Gb3 is likely due to less ligand, and not a reflection of reduced binding affinity of Stx1 to lyso-Gb3 (data not shown). While previous studies reported that Stx1 and Stx2 are able to bind to Lyso-Gb3 by thin layer chromatography
[309], receptor binding ELISA [310-312], or radio-labeled Stx [313], these studies did not compare binding of Lyso-Gb3 to native Gb3. Our results show weak binding of Stx to Lyso Gb3 when compared to native Gb3, which agrees with previous observations [309, 313].
Figure 4.7. Stx binding to Lyso- Gb3. Stx binding was assessed by ELISA at 10 nM for both Stx1 (A) and Stx2 (B) at 37⁰C. As negative controls, toxin was incubated in methanol-coated wells. The RFU signal is the mean of three independent experiments and error bars indicate SD. Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5.
103 | P a g e
4.6. Contribution of cholesterol to Stx binding
To determine whether Ch or PC is important to Stx2 binding, we assessed the binding in the absence of either Ch or PC (Figure 4.8). The absence of cholesterol caused a statistically significant decrease in the binding of Stx1 and Stx2. The presence of cholesterol alone caused a statistically significant increase in the binding of Stx2 to Gb3. These results are consistent with published data by other groups that demonstrate the presence of cholesterol modulates binding to glycosphingolipids [301, 314-317], and PC does not appear to be required for enhanced binding.
Figure 4.8. Comparison of Stx binding to Gb3 in absence of cholesterol or phosphatidylcholine. Stx binding was assessed by ELISA at 10 nM for both Stx1 and Stx2 at 37⁰C as described in Experimental Procedures. As negative controls, toxin was incubated in methanol, PC, Ch or PC+Ch coated wells. In all experiments, RFU values obtained in methanol were subtracted from each value in order to define a base level. The RFU signal is the mean of three independent experiments and error bars indicate SD. Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5.
Yahi et al. reported that cholesterol forms hydrogen bonds with glycosphingolipids by the interaction of the OH of cholesterol (donor group), the NH of sphingosine (acceptor group), and the oxygen atom of the glycosidic bond [acceptor group [315]]. These interactions change the glycolipid conformation and alter glycolipid interactions with proteins. For example, cholesterol has been reported to alter the ability of pathogens such as HIV to interact with the cell [318]. To investigate the role of the OH of cholesterol, we evaluated the binding of Stx1 and Stx2 to Gb3
104 | P a g e
in the presence of 5-α-Cholestane (5αCh) (Figure 4.9). This cholesterol analog lacks the OH group at Carbon 3 and has an alkane bond in Carbon 5 (Figure 4.9C). Stx1 bound equally well in the presence of Ch or 5αCh and PC (Figure 4.9A). In contrast, 5αCh failed to support binding of Stx2 in the presence of PC (Figure 4.9B). These results demonstrated that the presence of the OH group in cholesterol plays a role in modulating the binding of Stx2 but not Stx1.
Figure 4.9. Stx binding to Gb3 in presence of a cholesterol analog. Stx binding was assessed by ELISA at 10 nM for both Stx1 (A) and Stx2 (B) at 37⁰C. As negative controls, toxin was incubated in methanol-coated wells. The RFU signal is the mean of three independent experiments and error bars indicate SD. Statistical differences were calculated by the two-tailed Student's t-test using GraphPad Prism™ 5.
4.7. Stx cellular toxicity in vero protection assay
Little binding of Stx to glycolipids was observed at sub-nanomolar levels (Figures 4.3 and
4.5). However, cellular toxicity has been reported to occur at much lower concentrations [273].
Stx causes toxicity by cleaving the 28S rRNA of target cells, thereby inhibiting protein synthesis
[269, 270]. We assessed Stx-mediated inhibition of protein synthesis using Vero monkey kidney cells engineered to express a destabilized form of luciferase, Luc2P. Luc2P is targeted to the proteosome for degradation. Since it cannot accumulate in the cell, the amount of luciferase activity is proportional to the current rate of protein synthesis.
To assess the ability of glycolipids to neutralize cellular toxicity, serial dilutions of Stx were incubated in glycolipid-coated microtiter plates at 37º C for 1 hour, essentially as described in
Figure 4.5. The supernatant containing unbound toxin was transferred to plates containing the
105 | P a g e
Luc2P Vero cells. Protein synthesis inhibition was assessed after 4 hours of incubation with the toxin.
-11 In this assay, the ED50 for untreated Stx1 was 0.3 x 10 , and the ED50 for untreated Stx2 was 5 x 10-11. Pre-incubation of Stx1 in wells treated with methanol (Figure 4.10A, open inverted triangles) or PC+Ch (Figure 4.10A, open triangles) did not result in decreased toxicity, as seen by no change in ED50 compared to the untreated control (Figure 4.10A, insert). Pre- incubation with Gb4+PC+Ch (Figure 4.10A, open squares) was not able to protect Vero cells.
However, pre-incubation of Stx1 with Gb3+PC+Ch (Figure 4.10A, open circles) resulted in significantly reduced toxicity, with about a 10-fold increase in the ED50 compared to untreated
Stx1 (Figure 4.10A, insert). In contrast, Stx2 was not neutralized by any of the treatments since there were no significant differences in the ED50 values for treated or untreated toxin (Figure
4.10B: insert).
Figure 4.10. Vero protection studies. Stx cellular toxicity was assessed using luciferase activity of Luc2p Vero cells treated with dilutions of Stx1 (A) or Stx2 (B) pre- incubated with glycolipid mixtures as described in Figure 5. As negative controls, toxin was untreated or incubated in methanol-coated wells or PC+Ch. The results are the average of three independent experiments. Statistical difference was calculated between untreated control and Gb3+PC+Ch treatment by the two-tailed Student's t-test using GraphPad Prism™ 5 (***, P= 0.0002).
5. Discussion
The present study provides insights into the difference in receptor recognition by Stx1 and
Stx2. While Stx1 binds with similar affinity to the Pk glycan and the Gb3 glycolipid (Figure 4.5),
106 | P a g e
Stx2 does not recognize Pk alone, but can bind in the context of Gb3 glycolipid and other molecules (Figures 4-5). We found that Stx1 can also bind to the P tetrasaccharide (Figure 4.1) and the Gb4 glycolipid (Figure 4.4-4.5), which has not been reported previously. While Stx2 did not bind P glycan, it could bind to the glycolipid Gb4 (Figure 4.5). Given the differences in the ability of Stx1 and Stx2 to recognize glycan, it is intriguing that both toxins bind to the glycolipids
Gb3 and Gb4 with nearly identical affinity when PC and Ch are present (Figure 4.5).
Unlike Stx1, Stx2 binding to Gb3 is critically dependent on the presence of other compounds, either another glycolipid such as Gal-Cer or Ch (Figure 4.4). It is interesting to note that Stx2 but not Stx1 is associated with neurologic damage, and Gal-Cer is highly expressed on neuronal tissues [265, 319]. The second component could enhance binding of Stx2 to Gb3 either by directly contacting the toxin or by inducing Gb3 to assume a conformation more favorable for Stx2 binding. Ch has been shown to form hydrogen bonds with the ceramide on glycolipids, leading to conformational changes that make cells more susceptible to infection with
HIV [314-316]. We do not know if this mechanism is responsible for increased binding of Stx2 in the presence of Ch. However, we do not believe that the greatly improved binding of Stx2 to 1:1 mixture of Gb3 and Gal-Cer in the absence of Ch is achieved through conformational changes, since the ceramide of Gal-Cer and Gb3 is identical. An explanation that would account for the increased binding of Stx2 in the presence of Ch and Gal-Cer is that these molecules provide additional binding contacts. Stx2 could form hydrogen bonds with the galactose on Gal-Cer or with cholesterol.
In addition to the potential for additional binding contacts, the lower Hill coefficients observed for both Stx1 and Stx2 in the presence of cholesterol suggest that different classes of avidity are displayed on the plate surface, presumably due to heterogeneity in the distribution of the molecules. Inclusion of PC and Ch may favor formation of lipid microdomains that support
Stx binding to differing degrees, resulting in broadened binding curves due to overlapping ranges of avidity depending on the localized geometry of the glycans and how they interact with
107 | P a g e
the binding sites on the toxin. Clusters of glycolipids whose geometry precisely matches the binding sites within the toxin would allow maximum apparent affinity, and the reduced fluidity of the membrane upon addition of Ch would increase the lifetime of such localized glycolipid populations. This phenomenon may have important implications for the in vivo biological activity of the toxin, since such broadened binding curves exhibit detectable binding at very low toxin concentrations.
Studies with chimeric toxins where the Stx1 and Stx2 A- and B-subunits were reassorted demonstrated that potency tracks with the B-subunit of Stx2 [276, 278, 287], strongly suggesting that potency is determined by which cells are targeted, which is determined by receptor usage.
However, the current results do not explain the difference in potency of Stx1 and Stx2. An enormous disparity exists between the binding observed using biochemical assays compared to cellular susceptibility. The Kd values of Stx1 and Stx2 to Gb3 from this and previously published studies [276, 284] generally range between 10-7 M and 10-9 M. The concentration of toxin in blood at 50% lethal dose in mice is approximately 10-9 M for Stx1 and 10-11 M for Stx2 [273].
However, both Stx1 and Stx2 are toxic to primary human renal proximal tubular epithelial cells of
-13 the kidney with an ED50 of about 10 M [273] and to the Vero monkey kidney cell line with an
-11 ED50 of about 10 M (Figure 4.10). Since we are unable to observe any binding in vitro at these low doses, we examined the ability of toxin preincubated with glycolipid to protect Vero cells from Stx-mediated inhibition of protein synthesis. Even though nearly identical Kds were observed for Stx1 and Stx2 binding to Gb3 and Gb4 (Figure 4.4), Stx1 but not Stx2 was neutralized by preincubation with Gb3 mixed with PC+CH (Figure 4.10). These studies suggest that the in vitro glycolipid system replicates most of the elements need for cellular binding of
Stx1, but not Stx2.
Several properties of living cells could allow for toxin activity at concentrations where no binding occurs in biochemical systems. One major difference is the membranes of living cells are highly fluid and can form invaginations or protrusions, which cannot be formed by membrane
108 | P a g e
components bound to the rigid surface of microtiter plates. Stx1 has been shown to induce tubular membrane invaginations both in living cells and model membranes [296], and the high concentration of Stx1 in the tubules could drive toxin binding. Currently, there are no reports that
Stx2 can induce tubular invaginations.
In addition to membrane plasticity, living cells may express other molecules which bind
Stx1 or Stx2 with a higher affinity than Gb3. In the glycan array (Figure 4.2), Stx1 bound better to glycans containing GlcNAc at the third position instead of Glc (Galα1-4Galβ1-4GlcNAc versus Galα1-4Galβ1-4Glc). In other published reports [277, 285], Stx2 preferred a Pk mimic
(NAc-Pk: NAcGalα1-4Galβ1-4Glc) to native Pk. While these preferred glycans are not found on glycolipids, both are found on glycoproteins, and accumulating reports suggest that Stx may engage protein receptors. In 1999, Katagiri et al. were the first to report that Stx induced activation of tyrosine kinase within minutes of binding to a cell [320]. Recently, treatment with the B-pentamer from either Stx1 or Stx2 was shown to promote release of von Willebrand factor
(VWF) from endothelial cells [321] by a process that is dependent on Gb3 and cholesterol, and requires caveolin-1, but not clathrin, and Stx2B can initiate activation of the coagulation cascade in animal models of disease [322]. Furthermore, it has recently been shown that Stx1B and
Stx2B use different signaling pathways to promote VWF release [321] Activation of VWF release by Stx1B is associated with transient elevation of intracellular calcium, and requires both phospholipase C and protein kinase C. In contrast, activation of VWF release by Stx2B requires protein kinase A, which is activated in a cAMP-independent manner. Stx could activate a signaling pathway by binding to a protein receptor in a manner which mimics agonist activation.
Alternatively, Stx could promote receptor activation by a lectin-like mechanism. Lectins activate signaling pathways that respond to receptor-clustering. Like Stx, lectins possess multiple glycan- binding sites, and can crosslink receptors via N- or O-linked glycans present on receptor proteins. The presence of protein receptors could enhance Stx bind to cells. However, it is
109 | P a g e
important to recognize that living cells can internalize the toxin, and internalized toxin in a cellular system is equivalent to irreversible binding in a biochemical system.
Important questions regarding the pathogenesis of Stx-mediated disease remain unanswered. Why is Stx2 more likely to cause fatal disease than Stx1? Why are children more susceptible than adults? Is Stx-mediated killing of kidney epithelial cells more important than
Stx-mediated activation of the clotting cascade by kidney endothelial cells? Since hemolytic uremic syndrome patients who also display neurologic symptoms are more likely to succumb to fatal disease [319, 323], does Stx target the nervous system? Currently, only supportive care is available for patients with Stx-mediated disease. A detailed understanding of toxin binding preferences would allow us to identify the cells, organ systems, and even individuals that are most susceptible to the toxin. Such understanding is essential for the development of effective treatment strategies.
The results of this chapter have been published in:
Gallegos KM, Conrady DG, Karve SS, Gunasekera TS, Herr AB, et al. (2012)
Shiga Toxin Binding to Glycolipids and Glycans. PLoS ONE 7(2): e30368.
doi:10.1371/journal.pone.0030368
110 | P a g e
CHAPTER 5: FUTURE RESEARCH DIRECTIONS
Understanding invariant factors of pathogenic microbes is critical for the development of appropriate surveillance methods, preventive strategies and effective treatments. In this thesis, invariant determinants of two microbial pathogens, influenza virus and Shiga toxin of E. coli
O157:H7, were investigated to develop new diagnostics and potential treatment to replace traditional antibody recognition methods. First, we focused on the characterization of NA activity of whole viral particles as a marker for influenza detection (Chapter 2). Second, we studied the binding of influenza virus to host receptors to evaluate their potential as a capturing agents for viral detection (Chapter 3). Third, we analyze the binding of two different Shiga toxin forms to natural glycolipids and lipid mixtures in order to understand their difference in toxicity and affinity to host receptors to intend new strategies for treatment and diagnosis (Chapter 4).
Many questions concerning diagnosis of these two pathogens remain unanswered. From
Chapter 2, we found that by measuring NA activity in different conditions we can detect influenza NA and differentiate it from other pathogens such as parainfluenza and S. pneumoniae. From Chapter 3, we found that we can use host factors to capture influenza virus.
The future directions of research regarding the NA assay include its evaluation to detect influenza in patient samples such as nasal wash, nasal swabs, etc.
For initial studies with patient samples, we obtained nasal swabs from a broad epidemiological study for respiratory agents of Cincinnati Children's hospital (Figure 5.1). All patient samples were tested by Luminex mutliplex assay that analyzes of 17 different respiratory viral pathogens including influenza, parainfluenza, coronaviruses, Respiratory Syncytial Virus,
Human Metapneumovirus, Rhinovirus, and Adenovirus. Our initial results on the evaluation of the NA assay with patient samples (Figure 5.1) show that there is a range in which our assay is
100% specific (patient sample number 184, 169, 185, 116, 135, 134, 166, 168, and 138); however, there is also a range in which we are unable to detect influenza (patient sample
111 | P a g e
number 170, 177 and 194). It is possible to improve the sensitivity of the assay by maximizing the sample size. Each direct nasal specimen of a patient contains 200 to 500 μl of sample [324], which is diluted in 3000 μl of transport media [68, 325]; representing a 6 to 15 fold dilution. In this initial evaluation (Figure 5.1), samples were further diluted 3-fold making a total 12 to 45- fold dilution. It is possible that the sensitivity of the assay could be improved if we directly test
NA activity in patient specimens avoiding dilutions.
Figure 5.1. Detection of influenza in patient samples by NA activity assay. Patient were obtained following the standardized procedure approved by WHO [68] and tested for influenza using a multiplex PCR assay. The inclusion criteria of patients generally consisted of children under 13 years of age from the emergency department or hospitalized children who were admitted within 48 hours of enrollment. Children admission diagnoses included acute respiratory illness, apnea, asthma exacerbation, bronchiolitis, croup, cystic fibrosis exacerbation, syncytial virus febrile neonate, febrile seizure, influenza, fever with localizing signs, respiratory distress, pneumonia, pneumonitis, rule out sepsis, sinusitis, tonsillitis, pharyngitis strep throat, otitis media, and upper respiratory infection. Respiratory symptoms of emergency department hospitalized patients consisted of cough, nasal congestion, wheezing, sore throat, earache, dyspnea, and fever. Patient samples were tested blindly with the NA assay. Patient samples, virus-containing media (positive control) and virus free media (negative control) were diluted in 1:3 in PBS reaction buffer to a final concentration 2 mM CaCl2 and 10 μM MUNANA and were incubated for 3 H at 37°C and read as detailed in material and methods in Chapter 2. Where indicated, 10 mM final concentration Oseltamivir (Os) or 10 mM final concentration of Zanamivir (Za) was added to the reaction buffer. X-axis indicates the patient sample codes and the controls tested. The y- axis shows the measurement of the NA activity in RFU. NA activity of samples containing influenza virus A are shown in red, samples containing influenza virus type B are shown in green and negative sample are show as white bars.
112 | P a g e
Additionally to maximizing sample size, the sensitivity of the assay can be improved by using a more sensitive sialic acid substrate. NeuAc analogs include colorimetric or chemiluminescence substrates. Our initial studies with other sialic acid analogs include colorimetric substrates: 5-Bromo-4-chloro-3-indolyl α-D-N-acetylneuraminic (X-NeuAc, Figure
5.2) and 2-O-(p-Nitrophenyl)-α-D-N-acetylneuraminic acid (p-NP-NeuAc, Figure 5.3).
Figure 5.2. Evaluation of X-NeuAc substrate. Dilutions of in vitro culture virus were made in PBS pH 7.4 to a final concentration 0.63 mM of X- NeuAc and incubated overnight at 37°C. NA activity was detected by reading absorption at 640 mM. Data represents the results of a single experiment, n = 1.
Figure 5.3. Evaluation of p-NP- NeuAc substrate. Dilutions of in vitro culture virus and virus free media (negative control) were diluted in PBS pH 7 reaction buffer to a final concentration 0.1 mM CaCl2 and 1 mM p-NP-NeuAc and incubated for overnight at 37°C. NA activity was detected by reading absorption at 405 mM. Data represents the results of a single experiment, n = 1.
The limit of detection for NA activity assay with MUNANA fluorogenic substrate determined in Chapter 2 was 500-fold dilution. Our results with colorimetric NeuAc analogs show that these substrates are less sensitive than the flourogenic substrate (Chapter 2 and 3).
While X-NeuAc detects influenza virus up to 100-fold dilution, p-NP-NeuAc detects up to 270- dilution or less. In the case of X-NeuAc, the positive control absorbance (Figure 5.2, white bars) is only two fold different from the negative control (Figure 5.2, black bars). Similarly, our results with p-NP-NeuAc show that the positive control absorbance (Figure 5.3, white bars) is less than two fold different than the negative control (Figure 5.3, black bars). However, the data shown in
Figure 5.4 and 5.5 are the results of a single trial and the statistical significance was not
113 | P a g e
determined. Hence, more studies are need to further elucidate if these colorimetric substrates could be used in point-of-care test. To improve sensitivity of this NA activity assay, we suggest the use of more sensitive substrates such as a chemiluminescence sialic acid analog (e.g. NA-
Star of Applied Biosystems, USA).
To improve specificity of the assay and restrict false positive results, the reaction buffer could include non-viral NA inhibitors. We observed a range on false positive samples (patient sample number 171, 90, 92, and 172) in our NA assay (Figure 5.1). It is possible the NA activity of this negative sample came from bacterial co-infection. Since the admission of patients to the epidemiological study was broad and not specific for the evaluation of this influenza NA assay, it is possible that some influenza-negative samples include bacterial respiratory infection such as
S. pneumoniae; however, this needs further investigation. Future studies to improve this NA activity assay could include a bacterial NA inhibitor. Specific inhibitors for bacterial NA activity could be added to the reaction buffer. In this way, all NA activity from bacteria sources will be eliminated and NA activity detected will be from viral source only. Inhibitors for NanA, NA of S. pneumoniae, have been described previously; these include CHES [326] among others. In addition, to improve the specificity of this NA activity assay, we suggest to include a specific-viral capturing step, described as capturing assay in Chapter 3. From Chapter 3, we learned that host factors can be used to capture influenza virus. We found that fetuin and MBL were the best capturing factors evaluated in this study. If a capturing step specific for influenza virus is included in the NA assay, the specificity of the assay can be potentially improved. However, the sensitivity could be reduced if the capturing agents only bind to a small number of viral particles.
We also tested NA activity of patient swabs in the presence of influenza NA inhibitors
(NAI), Oseltamivir (10 mM) and Zanamivir (10 mM), shown in Figure 5.1 as doted bars and black bar, respectively. We observed that most or the influenza positive samples had little to no
NA activity in the presence of Oseltamivir. Only patient sample 172 showed significantly high NA activity in the presence of this high concentration of Oseltamivir (10 mM). Since Ki levels of
114 | P a g e
current influenza strains are reported to be on the nanoMolar range [327], we believe that the
NA activity observed in patient sample 172 in Oseltamivir, might not be from influenza source but could be bacterial origin. In order to be able to differentiate NA inhibitor-resistant or sensitive influenza strains, we believe that bioavailable concentration of NAI should be used for this assay. Oseltamivir treated patients have a blood concentration of 350 nM of the drug [236]; while Zanamivir treated patients have a nasal concentration of 30 nM of this inhibitor [328, 329].
Being able to distinguish from NA inhibitor-resistant or sensitive influenza strains in the NA assay will provide an advantage over current diagnostic test and will have a great impact on patient treatment. Hence, further study on the ability of NA assay to differentiate between NAI- sensitive and resistant influenza strains in the presence of NA inhibitors is needed.
Human influenza A viruses adapts quickly by the acquisition of mutations in all viral proteins or by interchange of genes with other strains limiting its recognition by antibodies. In addition, segmented genome and broad host spectrum (e.g. humans, horses, pigs, seals and birds) contribute to the appearance of new influenza strains that infect humans. Genetic reassortment regularly occurs between animal and human influenza strains, as well as between human strains, making it possible for new strains and variants to arise. Lately, human infections with a new variant influenza A H3N2 virus (H3N2v) have been reported in the USA since Fall
2011 [330]. We suggest that future research should include the evaluation of the NA assay to detect new variants of influenza virus.
Rapid tests, often termed point-of-care (POC) tests and are usually performed in less than an hour in the doctor's office. To make our NA activity assay a test more accessible in a doctor's office, we need to reduce reaction time, the use of lab equipment and steps. Current reaction time of the NA activity assay is 3H. The lab equipment needed to perform this test includes a
37°C-incubator and a fluorometer. Further studies should evaluate different reaction times and temperatures, and test the need of a pH change for reading fluorescence. Our studies in
Chapter 2 and 3 included a final pH change to 8 or 9 to detect fluorescence in samples.
115 | P a g e
Our initial studies to make this assay a point of care test included evaluation of NA assay at different reaction times and temperatures (Figure 5.4). Our results (Figure 5.4) show that the
NA assay can be performed a room temperature without significantly affecting sensitivity and avoiding the need of an incubator. To decide which reaction time was adequate for influenza detection, we determined the ratios of RFU of influenza NA activity over RFU of negative control
(Figure 5.5), to compare which conditions generate the biggest difference from its negative control. From Figure 5.5, we learned that reaction time can be decreased down to 90 minutes and still maintain good sensitivity. In addition, the readings at pH 6.5 (Figure 5.5, red lines) display equivalent differences from the negative control as the reading at pH 9 (Figure 5.5, blue lines).
Figure 5.4. Evaluation of different reaction times and temperatures. In vitro culture virus and virus-free media (negative control) were diluted in 1:50 in PBS pH 6.5 reaction buffer to a final concentration 0.1 mM CaCl2 and 10 μM MUNANA and incubated for 3 H at 37°C and room temperature (RT) and prior read as detail in material and methods in Chapter 2, final pH was changed to 9 (right panel) or not (left panel).
116 | P a g e
Even thought previous studies [331] and our studies in Chapter 2 indicate that fluorescence of cleaved MUNANA substrate is recommended to be read above pH 8, here we show that we can also avoid the change of final pH to read fluorescence since the differences from the negative control at pH 6.5 or pH 9 are equivalent (Figure 5.5). We believe that a change in the final pH change will be needed if the fluorometer used to read the sample is not sensitive enough to detect the final RFUs at pH 6.5 (Figure 5.4).
Figure 5.5. Relative NA activity of influenza strains comparing di- fferrent reaction times, incubation temperature and final pH. RFU values were obtained as detailed in Figure 5.4.
To sum up, we have demonstrated that NA activity assay is a good potential method of influenza detection and differentiation form other respiratory pathogens (Chapter 2). We also found binding of host receptors that show promise to potentially detect influenza virus. We also provide important evidence that direct binding occurs between influenza and mannose binding factors and EGFR or galectin. Further studies are required to fully elucidate the molecular mechanisms of that mediate binding and its relation to inflammatory responses upon influenza infection.
In Chapter 4, we found that Stx1 and Stx2 may bind to different residues in the Gb3 molecule and that the lipid environment in which Gb3 receptor is found modulates Shiga toxin binding; especially for Stx2. Variants of Stx include a total of seven members, which display different levels of toxicity in vivo and in vitro. Future studies should be performed in order to
117 | P a g e
understand if their differences in toxicity are also mediated by different residues in the Gb3 molecule or the Gb3 lipid environment.
The findings of this work advance not only the knowledge about influenza NA glycoprotein and binding of influenza viruses to host factors but also provide new perspectives in terms of development of potential treatment for Shiga toxin.
In conclusion, despite the progress in influenza diagnosis, since the first diagnostic test
(hemagglutination) took place about 70 years ago, the influenza diagnosis during recent 2009 pandemic was inaccurate delaying medication [77-79] which highlighted the continued importance of effective diagnosis and treatment. Influenza viruses is the major pathogen causing respiratory infections worldwide to date with a high impact on public health, however, we are making progress to developed a point-of-care test which is not based on antibody recognition.
118 | P a g e
REFERENCES
1. WHO, Media Centre of World Health Organization. Fact sheet N°211 Influenza, April
2009, World Health Organization: Geneva, Switzerland.
2. Thompson, W.W., Influenza-Associated Hospitalizations in the United States. JAMA : the
journal of the American Medical Association, 2004. 292(11): p. 1333-1340.
3. Ryan, K., et al., Sherris Medical Microbiology, 2010, McGraw-Hill: New York.
4. Treanor, J., Influenza viruses, including avian influence and swine influenza., in
Principles and Practice of Infectious Diseases, G. Mandell, J. Bennett, and R. Dolin,
Editors. 2009, Elsevier Churchill Livingstone: Philadelphia.
5. Turkington, C. and B. Ashby, The encyclopedia of infectious diseases ed. L.o.H.a. Living.
2007.
6. Cunha, B.A., Influenza: historical aspects of epidemics and pandemics. Infectious
disease clinics of North America, 2004. 18(1): p. 141-155.
7. Mandell, G.L., R.G. Douglas, and J.E. Bennett, Mandell, Douglas, and Bennett's
Principles and Practice of Infectious Diseases. 7th edition ed. Front Matter, ed. G.L.
Mandell, R.G. Douglas, and J.E. Bennett. Vol. 2. 2010, Philadelphia, PA: Churchill
Livingstone, An Imprint of Elsevier 2265-2288.
8. Mubareka, S. and P. Palese, Influenza Virus: The Biology of a Changing Virus Influenza
Vaccines for the Future, R. Rappuoli and G. Del Giudice, Editors. 2011, Springer Basel.
p. 3-26.
9. Nayak, D.P., Influenza virus morphogenesis and budding. Virus Research, 2009. 143(2):
p. 147.
10. Skehel, J.J. and D.C. Wiley, Receptor binding and membrane fusion in virus entry: The
Influenza Hemagglutinin. Annual Review of Biochemistry, 2000. 69(1): p. 531-569.
119 | P a g e
11. Kordyukova, L.V., et al., Site-specific attachment of palmitate or stearate to cytoplasmic
versus transmembrane cysteines is a common feature of viral spike proteins. Virology,
2010. 398(1): p. 49-56.
12. Rossman, J.S. and R.A. Lamb, Influenza virus assembly and budding. Virology, 2011.
411(2): p. 229-236.
13. Bertram, S., et al., Novel insights into proteolytic cleavage of influenza virus
hemagglutinin. Reviews in medical virology, 2010. 20(5): p. 298-310.
14. Suzuki, T., et al., Sialidase Activity of Influenza A Virus in an Endocytic Pathway
Enhances Viral Replication. Journal of virology, 2005. 79(18): p. 11705-11715.
15. McKimm-Breschkin, J., Resistance of influenza viruses to neuraminidase inhibitors — a
review. Antiviral Research, 2000. 47(1): p. 1-17.
16. Olsen, C.W., The emergence of novel swine influenza viruses in North America. Virus
Research, 2002. 85(2): p. 199-210.
17. Wu, W., Y.-H. Sun, and N. Pante, Nuclear import of influenza A viral ribonucleoprotein
complexes is mediated by two nuclear localization sequences on viral nucleoprotein.
Virology Journal, 2007. 4(1): p. 49.
18. Portela, A. and P. Digard, The influenza virus nucleoprotein: a multifunctional RNA-
binding protein pivotal to virus replication. Journal of general virology, 2002. 83(4): p.
723-734.
19. Ng, A.K.-L., et al., Structure of the influenza virus A H5N1 nucleoprotein: implications for
RNA binding, oligomerization, and vaccine design. The FASEB Journal, 2008. 22(10): p.
3638-3647.
20. Yuan, P., Crystal structure of an avian influenza polymerase PAN reveals an
endonuclease active site. Nature (London), 2009. 458(7240): p. 909-913.
21. Gonzalez, S., Distinct regions of influenza virus PB1 polymerase subunit recognize
vRNA and cRNA templates. The EMBO journal, 1999. 18(13): p. 3767-3775.
120 | P a g e
22. Robb, N.C., NS2/NEP protein regulates transcription and replication of the influenza
virus RNA genome. Journal of general virology, 2009. 90(6): p. 1398-1407.
23. Imai, H., et al., The HA and NS Genes of Human H5N1 Influenza A Virus Contribute to
High Virulence in Ferrets. PLoS Pathog, 2010. 6(9): p. e1001106.
24. Danckwardt, S., 3′ end mRNA processing: molecular mechanisms and implications for
health and disease. The EMBO journal, 2008. 27(3): p. 482-498.
25. Hale, B.G., The multifunctional NS1 protein of influenza A viruses. Journal of general
virology, 2008. 89(10): p. 2359-2376.
26. Ito, T., et al., Evolutionary analysis of the influenza A virus M gene with comparison of
the M1 and M2 proteins. Journal of virology, 1991. 65(10): p. 5491-5498.
27. Reid, A.H., et al., Characterization of the 1918 “Spanish” Influenza Virus Matrix Gene
Segment. Journal of virology, 2002. 76(21): p. 10717-10723.
28. Gomez-Puertas, P., Influenza Virus Matrix Protein Is the Major Driving Force in Virus
Budding. Journal of virology, 2000. 74(24): p. 11538-11547.
29. Hu, F., Mechanisms of Proton Conduction and Gating in Influenza M2 Proton Channels
from Solid-State NMR. Science (New York, N.Y.), 2010. 330(6003): p. 505-508.
30. Whittaker, G., Intracellular trafficking of influenza virus: clinical implications for molecular
medicine. Expert reviews in molecular medicine, 2001. 2001: p. 1-13.
31. Lakadamyali, M., M.J. Rust, and X. Zhuang, Endocytosis of influenza viruses. Microbes
and infection, 2004. 6(10): p. 929-936.
32. Eierhoff, T., The epidermal growth factor receptor (EGFR) promotes uptake of influenza
A viruses (IAV) into host cells. PLoS pathogens, 2010. 6(9): p. e1001099.
33. Nanbo, A., et al., Ebolavirus Is Internalized into Host Cells via Macropinocytosis in a Viral
Glycoprotein-Dependent Manner. PLoS Pathog, 2010. 6(9): p. e1001121.
34. Harrison, S.C., Viral membrane fusion. Nature structural & molecular biology, 2008.
15(7): p. 690.
121 | P a g e
35. Morizono, K. and I.S.Y. Chen, Receptors and tropisms of envelope viruses. Current
Opinion in Virology, 2011. 1(1): p. 13-18.
36. Han, X., Membrane structure and fusion-triggering conformational change of the fusion
domain from influenza hemagglutinin. Nature structural biology, 2001. 8(8): p. 715-720.
37. Lee, K.K., Architecture of a nascent viral fusion pore. EMBO J, 2010. 29(7): p. 1299-
1311.
38. Betakova, T., M2 Protein-A Proton Channel of Influenza A Virus. Current Pharmaceutical
Design, 2007. 13(31): p. 3231-3235.
39. Stouffer, A.L., et al., Structural basis for the function and inhibition of an influenza virus
proton channel. Nature, 2008. 451(7178): p. 596-599.
40. Moscona, A., Medical Management of Influenza Infection. Annual review of medicine,
2008. 59(1): p. 397-413.
41. Dolamore, M., Influenza prophylaxis in the long-term care facility: a case-control study of
the risk factors for adverse drug reactions to amantadine. Current Therapeutic Research,
2003. 64(9): p. 753-763.
42. Pizzorno, A., Influenza Drug Resistance. Seminars in respiratory and critical care
medicine, 2011. 32(04): p. 409.
43. Wang, R. and J.K. Taubenberger, Methods for molecular surveillance of influenza.
Expert Review of Anti-infective Therapy, 2010. 8(5): p. 517-527.
44. Update: Drug Susceptibility of Swine-Origin Influenza A (H1N1) Viruses, April 2009.
JAMA: The Journal of the American Medical Association, 2009. 301(20): p. 2086-2087.
45. Yu, L. Chinese Influenza Weekly Report week 45/2011. Chinese Influenza Weekly
Report, 2011. 6.
46. De Clercq, E., Antiviral agents active against influenza A viruses. Nat Rev Drug Discov,
2006. 5(12): p. 1015-1025.
122 | P a g e
47. Wang, P., P. Palese, and R.E. O'Neill, The NPI-1/NPI-3 (karyopherin alpha) binding site
on the influenza a virus nucleoprotein NP is a nonconventional nuclear localization
signal. Journal of virology, 1997. 71(3): p. 1850-6.
48. Engelhardt, O.G., Functional association between viral and cellular transcription during
influenza virus infection. Reviews in medical virology, 2006. 16(5): p. 329-345.
49. Webster, R.G., et al., Evolution and ecology of influenza A viruses. Microbiological
Reviews, 1992. 56(1): p. 152-179.
50. Jorba, N., E. Area, and J. Ortín, Oligomerization of the influenza virus polymerase
complex in vivo. Journal of general virology, 2008. 89(2): p. 520-524.
51. Digard, P., L. Tiley, and D. Elton, Orthomyxovirus Genome Transcription and Replication
Viral Genome Replication, K.D. Raney, M. Gotte, and C.E. Cameron, Editors. 2009,
Springer US. p. 163-180.
52. Read, E.K.C. and P. Digard, Individual influenza A virus mRNAs show differential
dependence on cellular NXF1/TAP for their nuclear export. Journal of general virology,
2010. 91(5): p. 1290-1301.
53. Wu, C.-Y., K.-S. Jeng, and M.M.-C. Lai, The SUMOylation of Matrix Protein M1
Modulates the Assembly and Morphogenesis of Influenza A Virus. Journal of virology,
2011. 85(13): p. 6618-6628.
54. David, R., Virology: Severing the bud. Nature reviews. Microbiology, 2010. 8(11): p. 757-
757.
55. Hutchinson, E.C., et al., Genome packaging in influenza A virus. Journal of general
virology, 2010. 91(2): p. 313-328.
56. Rossman, J.S., et al., Influenza Virus M2 Protein Mediates ESCRT-Independent
Membrane Scission. Cell, 2010. 142(6): p. 902-913.
57. Shaw, M.L., Cellular proteins in influenza virus particles. PLoS pathogens, 2008. 4(6): p.
e1000085.
123 | P a g e
58. Min, J.-Y., Cellular targets for influenza drugs. Nature biotechnology, 2010. 28(3): p. 239-
240.
59. Wu, S., J.P. Metcalf, and W. Wu, Innate immune response to influenza virus. Current
Opinion in Infectious Diseases, 2011. 24(3): p. 235-240
10.1097/QCO.0b013e328344c0e3.
60. Peiris, J.S.M., K.P.Y. Hui, and H.-L. Yen, Host response to influenza virus: protection
versus immunopathology. Current Opinion in Immunology, 2010. 22(4): p. 475-481.
61. Yount, J.S., et al., Palmitoylome profiling reveals S-palmitoylation–dependent antiviral
activity of IFITM3. Nat Chem Biol, 2010. 6(8): p. 610-614.
62. Di Domizio, J., et al., TLR7 stimulation in human plasmacytoid dendritic cells leads to the
induction of early IFN-inducible genes in the absence of type I IFN. Blood, 2009. 114(9):
p. 1794-1802.
63. Le Goffic, R., et al., Cutting Edge: Influenza A Virus Activates TLR3-Dependent
Inflammatory and RIG-I-Dependent Antiviral Responses in Human Lung Epithelial Cells.
The Journal of Immunology, 2007. 178(6): p. 3368-3372.
64. Pichlmair, A., et al., RIG-I-Mediated Antiviral Responses to Single-Stranded RNA
Bearing 5'-Phosphates. Science, 2006. 314(5801): p. 997-1001.
65. Allen, I.C., et al., The NLRP3 Inflammasome Mediates In Vivo Innate Immunity to
Influenza A Virus through Recognition of Viral RNA. Immunity, 2009. 30(4): p. 556-565.
66. Thomas, P.G., et al., The Intracellular Sensor NLRP3 Mediates Key Innate and Healing
Responses to Influenza A Virus via the Regulation of Caspase-1. Immunity, 2009. 30(4):
p. 566-575.
67. Ichinohe, T., et al., Inflammasome recognition of influenza virus is essential for adaptive
immune responses. The Journal of Experimental Medicine, 2009. 206(1): p. 79-87.
68. WHO, Manual for the laboratory diagnosis and virological surveillance of influenza, 2011,
World Health Organization Press: Geneva, Switzerland.
124 | P a g e
69. Yewdell, J., Viva la revolución: rethinking influenza A virus antigenic drift. Current
Opinion in Virology, 2011. 1(3): p. 177-183.
70. Landry, M.L., Diagnostic tests for influenza infection. Current opinion in pediatrics, 2011.
23(1): p. 91-97.
71. Uyeki, Timothy M., Low Sensitivity of Rapid Diagnostic Test for Influenza. Clinical
Infectious Diseases, 2009. 48(9): p. e89-e92.
72. WHO, Use of influenza rapid diagnostic tests, W.H.O. Press, Editor 2010, World Health
Organization on behalf of the Special Programme for Research and Training in Tropical
Diseases 2010: Geneva, Switzerland. p. 27.
73. Gavin, P.J. and R.B. Thomson Jr, Review of Rapid Diagnostic Tests for Influenza.
Clinical and Applied Immunology Reviews, 2004. 4(3): p. 151-172.
74. Katz, J.M., K. Hancock, and X. Xu, Serologic assays for influenza surveillance, diagnosis
and vaccine evaluation. Expert Review of Anti-infective Therapy, 2011. 9(6): p. 669-683.
75. Hurt, A.C., et al., Performance of six influenza rapid tests in detecting human influenza in
clinical specimens. Journal of Clinical Virology, 2007. 39(2): p. 132-135.
76. CDC Seasonal Influenza (Flu) website at CDC Centers for Disease Control and
Prevention, National Center for Immunization and Respiratory Diseases (NCIRD), 2012.
77. Kao, T.M., Rapid Diagnostic Tests and Severity of Illness in Pandemic (H1N1) 2009,
Taiwan. Emerging infectious diseases, 2010. 16(7): p. 1181.
78. Sakai-Tagawa, Y., et al., Sensitivity of Influenza Rapid Diagnostic Tests to H5N1 and
2009 Pandemic H1N1 Viruses. Journal of clinical microbiology, 2010. 48(8): p. 2872-
2877.
79. Drexler, J.F., Poor Clinical Sensitivity of Rapid Antigen Test for Influenza A Pandemic
(H1N1) 2009 Virus. Emerging infectious diseases, 2009. 15(10): p. 1662.
80. Diggle, M. and S. Clarke, Pyrosequencing™. Molecular Biotechnology, 2004. 28(2): p.
129-137.
125 | P a g e
81. Deyde, V.M., et al., Detection of Molecular Markers of Antiviral Resistance in Influenza A
(H5N1) Viruses Using a Pyrosequencing Method. Antimicrobial agents and
chemotherapy, 2009. 53(3): p. 1039-1047.
82. Deyde, V.M., et al., Pyrosequencing as a tool to detect molecular markers of resistance
to neuraminidase inhibitors in seasonal influenza A viruses. Antiviral Research, 2009.
81(1): p. 16-24.
83. Deng, Y.-M., N. Caldwell, and I.G. Barr, Rapid Detection and Subtyping of Human
Influenza A Viruses and Reassortants by Pyrosequencing. PLoS ONE, 2011. 6(8): p.
e23400.
84. Weis, W., et al., Structure of the influenza virus haemagglutinin complexed with its
receptor, sialic acid. Nature, 1988. 333(6172): p. 426-431.
85. Gamblin, S.J., Influenza Hemagglutinin and Neuraminidase Membrane Glycoproteins.
The Journal of biological chemistry, 2010. 285(37): p. 28403-28409.
86. Sauter, N.K., et al., Crystallographic detection of a second ligand binding site in influenza
virus hemagglutinin. Proceedings of the National Academy of Sciences, 1992. 89(1): p.
324-328.
87. Sauter, N., et al., Binding of influenza virus hemagglutinin to analogs of its cell-surface
receptor, sialic acid: analysis by proton nuclear magnetic resonance spectroscopy and
X-ray crystallography. Biochemistry, 1992. 31(40): p. 9609 - 9621.
88. Carroll, S., H. Higa, and J. Paulson, Different cell-surface receptor determinants of
antigenically similar influenza virus hemagglutinins. J Biol Chem, 1981. 256(16): p. 8357
- 8363.
89. Rogers, G. and J. Paulson, Receptor determinants of human and animal influenza virus
isolates: differences in receptor specificity of the H3 hemagglutinin based on species of
origin. Virology, 1983. 127(2): p. 361 - 373.
126 | P a g e
90. Gambaryan, A.S., et al., Receptor-binding properties of swine influenza viruses isolated
and propagated in MDCK cells. Virus Research, 2005. 114(1-2): p. 15-22.
91. Matrosovich, M., et al., Early alterations of the receptor-binding properties of H1, H2, and
H3 avian influenza virus hemagglutinins after their introduction into mammals. J Virol,
2000. 74(18): p. 8502 - 8512.
92. Rogers, G., et al., Single amino acid substitutions in influenza haemagglutinin change
receptor binding specificity. Nature, 1983. 304: p. 76 - 78.
93. Rogers, G., et al., Host-mediated selection of influenza receptor variants. Sialic acid
alpha 2,6Gal specific clones of A/duck/Ukraine/1/63 revert to sialic acid alpha 2,3Gal-
specific wild type in ovo. J Biol Chem, 1985. 260: p. 7362 - 7367.
94. Stevens, J., et al., Glycan microarray analysis of the hemagglutinins from modern and
pandemic influenza viruses reveals different receptor specificities. J Mol Biol, 2006.
355(5): p. 1143 - 1155.
95. Kumari, K., et al., Receptor binding specificity of recent human H3N2 influenza viruses.
Virology Journal, 2007. 4(1): p. 42.
96. Wu, W. and G. Air, Binding of influenza viruses to sialic acids: reassortant viruses with
A/NWS/33 hemagglutinin bind to alpha2,8-linked sialic acid. Virology, 2004. 325(2): p.
340 - 350.
97. Stevens, J., et al., Structure and receptor specificity of the hemagglutinin from an H5N1
influenza virus. Science, 2006. 312(5772): p. 404 - 410.
98. Yamada, S., et al., Haemagglutinin mutations responsible for the binding of H5N1
influenza A viruses to human-type receptors. Nature, 2006. 444(7117): p. 378 - 382.
99. van Riel, D., et al., Seasonal and Pandemic Human Influenza Viruses Attach Better to
Human Upper Respiratory Tract Epithelium than Avian Influenza Viruses. The American
Journal of Pathology, 2010. 176(4): p. 1614-1618.
127 | P a g e
100. Ibricevic, A., et al., Influenza Virus Receptor Specificity and Cell Tropism in Mouse and
Human Airway Epithelial Cells. Journal of virology, 2006. 80(15): p. 7469-7480.
101. Matrosovich, M.N., et al., Human and Avian Influenza Viruses Target Different Cell
Types in Cultures of Human Airway Epithelium. Proceedings of the National Academy of
Sciences of the United States of America, 2004. 101(13): p. 4620-4624.
102. Childs, R.A., et al., Correspondence: Receptor-binding specificity of pandemic influenza
A (H1N1) 2009 virus determined by carbohydrate microarray. Nat Biotech, 2009. 27(9):
p. 797-799.
103. Childs, R.A., et al., Receptor-binding specificity of pandemic influenza A (H1N1) 2009
virus determined by carbohydrate microarray. Nat Biotech, 2010. 28(2): p. 178-178.
104. Liao, H.-Y., et al., Differential Receptor Binding Affinities of Influenza Hemagglutinins on
Glycan Arrays. Journal of the American Chemical Society, 2010. 132(42): p. 14849-
14856.
105. Gulati, S., D. Smith, and G. Air, Diversity in the receptor binding specificity of influenza
H3 hemagglutinin, in Participating Investigators Meeting July 2011, NIH: Bethesda,
Maryland.
106. Viswanathan, K., et al., Glycans as receptors for influenza pathogenesis. Glycoconjugate
Journal, 2010. 27(6): p. 561-570.
107. Lewallen, D.M., Factors Affecting Protein-Glycan Specificity: Effect of Spacers and
Incubation Time. Chembiochem : a European journal of chemical biology, 2009. 10(9): p.
1486-1489.
108. Suzuki, Y., Sialobiology of Influenza: Molecular Mechanism of Host Range Variation of
Influenza Viruses. Biological & pharmaceutical bulletin, 2005. 28(3): p. 399-408.
109. Chen, J.-R., C. Ma, and C.-H. Wong, Vaccine design of hemagglutinin glycoprotein
against influenza. Trends in biotechnology, 2011. 29(9): p. 426-434.
128 | P a g e
110. Wei, C.-J., et al., Cross-Neutralization of 1918 and 2009 Influenza Viruses: Role of
Glycans in Viral Evolution and Vaccine Design. Science Translational Medicine, 2010.
2(24): p. 24ra21.
111. Wang, C.-C., et al., Glycans on influenza hemagglutinin affect receptor binding and
immune response. Proceedings of the National Academy of Sciences, 2009. 106(43): p.
18137-18142.
112. Varki, A., Glycan-based interactions involving vertebrate sialic-acid-recognizing proteins.
Nature, 2007. 446(7139): p. 1023-1029.
113. Vigerust, D.J., Virus glycosylation: role in virulence and immune interactions. Trends in
microbiology (Regular ed.), 2007. 15(5): p. 211.
114. Das, S.R., PNAS Plus: Fitness costs limit influenza A virus hemagglutinin glycosylation
as an immune evasion strategy. PNAS : Proceedings of the National Academy of
Sciences, 2011. 108(51): p. E1417-E1422.
115. Ramos, I., et al., Effects of receptor binding specificity of avian influenza virus on the
human innate immune response. Journal of virology, 2011.
116. Oshansky, C.M., et al., Avian Influenza Viruses Infect Primary Human Bronchial
Epithelial Cells Unconstrained by Sialic Acid α2,3 Residues. PLoS ONE, 2011. 6(6): p.
e21183.
117. Takasaki, S., Asparagine-linked sugar chains of fetuin: occurrence of tetrasialyl
triantennary sugar chains containing the Gal.beta.1 .fwdarw. 3GlcNAc sequence.
Biochemistry (Easton), 1986. 25(19): p. 5709-5715.
118. Marinina, V.P., The Effect of Losing Glycosylation Sites Near the Receptor-Binding
Region on the Receptor Phenotype of the Human Influenza Virus H1N1. Molecular
biology (New York), 2003. 37(3): p. 468-472.
129 | P a g e
119. Bhatia, A. and R. Kast, How influenza’s neuraminidase promotes virulence and creates
localized lung mucosa immunodeficiency. Cellular & Molecular Biology Letters, 2007.
12(1): p. 111-119.
120. Gong, J., W. Xu, and J. Zhang, Structure and functions of influenza virus neuraminidase
Current Medicinal Chemistry, 2007. 14: p. 113-122.
121. Garman, E., The Structure, Function, and Inhibition of Influenza Virus Neuraminidase,
2005, Kluwer Academic/Plenum Publishers. p. 247-267.
122. Yen, H.L., Importance of Neuraminidase Active-Site Residues to the Neuraminidase
Inhibitor Resistance of Influenza Viruses. Journal of virology, 2006. 80(17): p. 8787-
8795.
123. Taylor, N.R. and M. von Itzstein, Molecular Modeling Studies on Ligand Binding to
Sialidase from Influenza Virus and the Mechanism of Catalysis. Journal of Medicinal
Chemistry, 1994. 37(5): p. 616-624.
124. Burmeister, W.P., The 2.2 A resolution crystal structure of influenza B neuraminidase
and its complex with sialic acid. The EMBO journal, 1992. 11(1): p. 49.
125. Ohuchi, M., Roles of neuraminidase in the initial stage of influenza virus infection.
Microbes and infection, 2006. 8(5): p. 1287.
126. Matrosovich, M.N., Neuraminidase Is Important for the Initiation of Influenza Virus
Infection in Human Airway Epithelium. Journal of virology, 2004. 78(22): p. 12665-12667.
127. Burmeister, W.P., Calcium is needed for the thermostability of influenza B virus
neuraminidase. Journal of general virology, 1994. 75: p. 381.
128. Carroll, S.M. and J.C. Paulson, Complete metal ion requirement of influenza virus N1
neuraminidases. Archives of Virology, 1982. 71(3): p. 273-277.
129. Kiyotani, K., Enzymological characteristics of avian influenza A virus neuraminidase.
Microbiology and immunology, 1987. 31(11): p. 1131-5.
130 | P a g e
130. Chong, A.K.J., M.S. Pegg, and M. von Itzstein, Influenza virus sialidase: effect of calcium
on steady-state kinetic parameters. Biochimica et Biophysica Acta (BBA) - Protein
Structure and Molecular Enzymology, 1991. 1077(1): p. 65-71.
131. Varghese, J.N., Three-dimensional structure of the neuraminidase of influenza virus
A/Tokyo/3/67 at 2· 2 Å resolution. Journal of molecular biology, 1991. 221(2): p. 473.
132. Li, Q., et al., The 2009 pandemic H1N1 neuraminidase N1 lacks the 150-cavity in its
active site. Nat Struct Mol Biol, 2010. 17(10): p. 1266-1268.
133. Russell, R.J., et al., The structure of H5N1 avian influenza neuraminidase suggests new
opportunities for drug design. Nature, 2006. 443(7107): p. 45-49.
134. Xu, X., et al., Structural Characterization of the 1918 Influenza Virus H1N1
Neuraminidase. Journal of virology, 2008. 82(21): p. 10493-10501.
135. Johansson, B.E., Variation in the Divalent Cation Requirements of Influenza A Virus N2
Neuraminidases. Journal of biochemistry (Tokyo), 2003. 134(3): p. 345-352.
136. Smith, B.J., et al., Structure of a calcium-deficient form of influenza virus neuraminidase:
implications for substrate binding. Acta Crystallographica Section D, 2006. 62(9): p. 947-
952.
137. Lawrenz, M., Impact of calcium on N1 influenza neuraminidase dynamics and binding
free energy. Proteins, structure, function, and bioinformatics, 2010. 78(11): p. n-a-n/a.
138. Alymova, I.V., G. Taylor, and A. Portner, Neuraminidase Inhibitors as Antiviral Agents.
Current Drug Targets - Infectious Disorders, 2005. 5(4): p. 401-409.
139. Colman, P.M., New Antivirals and Drug Resistance. Annual Review of Biochemistry,
2009. 78(1): p. 95-118.
140. Gubareva, L.V., Molecular mechanisms of influenza virus resistance to neuraminidase
inhibitors. Virus Research, 2004. 103(1-2): p. 199.
141. von Itzstein, M., The war against influenza: discovery and development of sialidase
inhibitors. Nat Rev Drug Discov, 2007. 6(12): p. 967-974.
131 | P a g e
142. Wang, M., Influenza A Virus N5 Neuraminidase Has an Extended 150-Cavity. Journal of
virology, 2011. 85(16): p. 8431-8435.
143. Russell, R.J., The structure of H5N1 avian influenza neuraminidase suggests new
opportunities for drug design. Nature (London), 2006. 443(7107): p. 45-49.
144. Stoll, V., Influenza Neuraminidase Inhibitors: Structure-Based Design of a Novel
Inhibitor Series. Biochemistry (Easton), 2003. 42(3): p. 718-727.
145. Ferraris, O., N. Kessler, and B. Lina, Sensitivity of influenza viruses to zanamivir and
oseltamivir: A study performed on viruses circulating in France prior to the introduction of
neuraminidase inhibitors in clinical practice. Antiviral Research, 2005. 68(1): p. 43-48.
146. Collins, P.J., Crystal structures of oseltamivir-resistant influenza virus neuraminidase
mutants. Nature (London), 2008. 453(7199): p. 1258-1261.
147. Nguyen, H.T., Assessment of Pandemic and Seasonal Influenza A (H1N1) Virus
Susceptibility to Neuraminidase Inhibitors in Three Enzyme Activity Inhibition Assays.
Antimicrobial agents and chemotherapy, 2010. 54(9): p. 3671-3677.
148. Tramontana AR, et al., Oseltamivir resistance in adult oncology and hematology patients
infected with pandemic (H1N1) 2009 virus, Australia. Emerg Infect Dis, 2010. 16(7): p.
1068-1075.
149. Abed, Y., et al., Role of Permissive Neuraminidase Mutations in Influenza
A/Brisbane/59/2007-like (H1N1) Viruses. PLoS Pathog, 2011. 7(12): p. e1002431.
150. Hatakeyama, S., et al., Emergence of Influenza B Viruses With Reduced Sensitivity to
Neuraminidase Inhibitors. JAMA: The Journal of the American Medical Association,
2007. 297(13): p. 1435-1442.
151. Sleeman, K., et al., Influenza B viruses with mutation in the neuraminidase active site,
North Carolina, USA, 2010−11. Emerg Infect Dis, 2011. 17(11): p. 2043-2046.
132 | P a g e
152. Gubareva, Larisa V., Evidence for Zanamivir Resistance in an Immunocompromised
Child Infected with Influenza B Virus. The Journal of infectious diseases, 1998. 178(5): p.
1257-1262.
153. Alberts B, et al., eds. Molecular Biology of the Cell,. 4th edition ed. 2002, Garland
Science: New York:. Chapter 13: Intracellular Vesicular Traffic.
154. Brass, A.L., The IFITM proteins mediate cellular resistance to influenza A H1N1 virus,
West Nile virus, and dengue virus. Cell (Cambridge), 2009. 139(7): p. 1243.
155. Keil, W., Carbohydrates of influenza virus. Structural elucidation of the individual glycans
of the FPV hemagglutinin by two-dimensional 1H nmr and methylation analysis. The
EMBO journal, 1985. 4(10): p. 2711.
156. Roberts, P.C., Role of conserved glycosylation sites in maturation and transport of
influenza A virus hemagglutinin. Journal of virology, 1993. 67(6): p. 3048.
157. Stahl, P.D. and R.A.B. Ezekowitz, The mannose receptor is a pattern recognition
receptor involved in host defense. Current Opinion in Immunology, 1998. 10(1): p. 50-55.
158. Bowden, T.A. and E.E. Fry, Role of Glycoproteins in Virus–Human Cell Interactions
Functional and Structural Proteomics of Glycoproteins, R. Owens and J. Nettleship,
Editors. 2011, Springer Netherlands. p. 159-180.
159. Kerrigan, A.M. and G.D. Brown, C-type lectins and phagocytosis. Immunobiology, 2009.
214(7): p. 562-575.
160. Wu, L. and V.N. KewalRamani, Dendritic-cell interactions with HIV: infection and viral
dissemination. Nat Rev Immunol, 2006. 6(11): p. 859-868.
161. Maeda, N., et al., The Cell Surface Receptor DC-SIGN Discriminates
betweenMycobacterium Species through Selective Recognition of the Mannose Caps on
Lipoarabinomannan. Journal of Biological Chemistry, 2003. 278(8): p. 5513-5516.
133 | P a g e
162. Lozach, P.-Y., et al., DC-SIGN and L-SIGN Are High Affinity Binding Receptors for
Hepatitis C Virus Glycoprotein E2. Journal of Biological Chemistry, 2003. 278(22): p.
20358-20366.
163. Cambi, A., et al., Dendritic Cell Interaction with Candida albicans Critically Depends on
N-Linked Mannan. Journal of Biological Chemistry, 2008. 283(29): p. 20590-20599.
164. Zhang, P., et al., Human dendritic cell-specific intercellular adhesion molecule-grabbing
nonintegrin (CD209) is a receptor for Yersinia pestis that promotes phagocytosis by
dendritic cells. Infect Immun, 2008. 76(5): p. 2070-2079.
165. Miller, J.L., et al., The Mannose Receptor Mediates Dengue Virus Infection of
Macrophages. PLoS Pathog, 2008. 4(2): p. e17.
166. Cormier, E.G., et al., L-SIGN (CD209L) and DC-SIGN (CD209) mediate transinfection of
liver cells by hepatitis C virus. Proceedings of the National Academy of Sciences of the
United States of America, 2004. 101(39): p. 14067-14072.
167. Bergman, M.P., et al., Helicobacter pylori Modulates the T Helper Cell 1/T Helper Cell 2
Balance through Phase-variable Interaction between Lipopolysaccharide and DC-SIGN.
The Journal of Experimental Medicine, 2004. 200(8): p. 979-990.
168. Feinberg, H., et al., Multiple Modes of Binding Enhance the Affinity of DC-SIGN for High
Mannose N-Linked Glycans Found on Viral Glycoproteins. Journal of Biological
Chemistry, 2007. 282(6): p. 4202-4209.
169. Mitchell, D.A., A.J. Fadden, and K. Drickamer, A Novel Mechanism of Carbohydrate
Recognition by the C-type Lectins DC-SIGN and DC-SIGNR. Journal of Biological
Chemistry, 2001. 276(31): p. 28939-28945.
170. Wang, S.-F., et al., DC-SIGN mediates avian H5N1 influenza virus infection in cis and in
trans. Biochemical and Biophysical Research Communications, 2008. 373(4): p. 561-
566.
134 | P a g e
171. Londrigan, S.L., et al., N-Linked Glycosylation Facilitates Sialic Acid-Independent
Attachment and Entry of Influenza A Viruses into Cells Expressing DC-SIGN or L-SIGN.
Journal of virology, 2011. 85(6): p. 2990-3000.
172. Sato, Y., et al., High mannose-specific lectin (KAA-2) from the red alga Kappaphycus
alvarezii potently inhibits influenza virus infection in a strain-independent manner.
Biochemical and Biophysical Research Communications, 2011. 405(2): p. 291-296.
173. Sato, Y., et al., High Mannose-binding Lectin with Preference for the Cluster of α1–2-
Mannose from the Green Alga Boodlea coacta Is a Potent Entry Inhibitor of HIV-1 and
Influenza Viruses. Journal of Biological Chemistry, 2011. 286(22): p. 19446-19458.
174. Reading, P.C., J.L. Miller, and E.M. Anders, Involvement of the Mannose Receptor in
Infection of Macrophages by Influenza Virus. Journal of virology, 2000. 74(11): p. 5190-
5197.
175. Man, L.-T., Thesis: The Role of Mannose Binding Lectin in Influenza Virus Infection, in
Department of Pediatrics and Adolescent Medicine2009, The University of Hong Kong:
Hong Kong. p. 155.
176. Upham, J.P., et al., Macrophage Receptors for Influenza A Virus: Role of the
Macrophage Galactose-Type Lectin and Mannose Receptor in Viral Entry. Journal of
virology, 2010. 84(8): p. 3730-3737.
177. Eisen, D.P., et al., No association between mannose-binding lectin deficiency and H1N1
2009 infection observed during the first season of this novel pandemic influenza virus.
Human Immunology, 2011. 72(11): p. 1091-1094.
178. Ling, M.T., et al., Mannose-Binding Lectin Contributes to Deleterious Inflammatory
Response in Pandemic H1N1 and Avian H9N2 Infection. Journal of Infectious Diseases,
2012. 205(1): p. 44-53.
135 | P a g e
179. Chang, W.-C., et al., Lack of the pattern recognition molecule mannose-binding lectin
increases susceptibility to influenza A virus infection. BMC Immunology, 2010. 11(1): p.
64.
180. Pohlmann, S., DC-SIGNR, a DC-SIGN homologue expressed in endothelial cells, binds
to human and simian immunodeficiency viruses and activates infection in trans. PNAS :
Proceedings of the National Academy of Sciences, 2001. 98(5): p. 2670-2675.
181. Srivastava, V., et al., Toll-like Receptor 2 and DC-SIGNR1 Differentially Regulate
Suppressors of Cytokine Signaling 1 in Dendritic Cells during Mycobacterium
tuberculosis Infection. Journal of Biological Chemistry, 2009. 284(38): p. 25532-25541.
182. Sugita, H., Neuraminidase-producing ability of intestinal bacteria isolated from coastal
fish. Letters in applied microbiology, 2000. 31(1): p. 10-13.
183. Moncla, B.J., Sialidase (neuraminidase) activity among gram-negative anaerobic and
capnophilic bacteria. Journal of clinical microbiology, 1990. 28(3): p. 422.
184. Parker, D., The NanA Neuraminidase of Streptococcus pneumoniae Is Involved in
Biofilm Formation. Infection and immunity, 2009. 77(9): p. 3722-3730.
185. McGreal, E.P., The carbohydrate-recognition domain of Dectin-2 is a C-type lectin with
specificity for high mannose. Glycobiology (Oxford), 2006. 16(5): p. 422-430.
186. Yuan, P., Structure of the Newcastle disease virus hemagglutinin-neuraminidase (HN)
ectodomain reveals a four-helix bundle stalk. PNAS : Proceedings of the National
Academy of Sciences, 2011. 108(36): p. 14920-14925.
187. Moscona, A., Entry of parainfluenza virus into cells as a target for interrupting childhood
respiratory disease. The Journal of clinical investigation, 2005. 115(7): p. 1688.
188. Hughes, A.K., et al., Cytotoxic effect of Shiga toxin-1 on human glomerular epithelial
cells1. Kidney Int, 2000. 57(6): p. 2350-2359.
136 | P a g e
189. Alymova, I.V., Efficacy of Novel Hemagglutinin-Neuraminidase Inhibitors BCX 2798 and
BCX 2855 against Human Parainfluenza Viruses In Vitro and In Vivo. Antimicrobial
agents and chemotherapy, 2004. 48(5): p. 1495-1502.
190. Alymova, I.V., Receptor-Binding Specificity of the Human Parainfluenza Virus Type 1
Hemagglutinin-Neuraminidase Glycoprotein. Glycobiology (Oxford), 2011.
191. Amonsen, M., et al., Human parainfluenza viruses hPIV1 and hPIV3 bind
oligosaccharides with alpha2-3-linked sialic acids that are distinct from those bound by
H5 avian influenza virus hemagglutinin. Journal of virology, 2007. 81(15): p. 8341-5.
192. Lawrence, M.C., Structure of the haemagglutinin-neuraminidase from human
parainfluenza virus type III. Journal of molecular biology, 2004. 335(5): p. 1343.
193. Hata, K., Limited Inhibitory Effects of Oseltamivir and Zanamivir on Human Sialidases.
Antimicrobial agents and chemotherapy, 2008. 52(10): p. 3484-3491.
194. Porotto, M., Inhibition of Parainfluenza Virus Type 3 and Newcastle Disease Virus
Hemagglutinin-Neuraminidase Receptor Binding: Effect of Receptor Avidity and Steric
Hindrance at the Inhibitor Binding Sites. Journal of virology, 2004. 78(24): p. 13911-
13919.
195. Vlasak, R., Human and bovine coronaviruses recognize sialic acid-containing receptors
similar to those of influenza C viruses. PNAS : Proceedings of the National Academy of
Sciences, 1988. 85(12): p. 4526.
196. Zeng, Q., Structure of coronavirus hemagglutinin-esterase offers insight into corona and
influenza virus evolution. PNAS : Proceedings of the National Academy of Sciences,
2008. 105(26): p. 9065-9069.
197. Schwegmann-Weßels, C., Sialic acids as receptor determinants for coronaviruses.
Glycoconjugate Journal, 2006. 23(1-2): p. 51-58.
198. Tan, E.L., Inhibition of SARS coronavirus infection in vitro with clinically approved
antiviral drugs. Emerging infectious diseases, 2004. 10(4): p. 581.
137 | P a g e
199. Vimr, E., To sialylate, or not to sialylate: that is the question. Trends in microbiology
(Regular ed.), 2002. 10(6): p. 254-257.
200. Mishin, V.P., et al., N-Linked Glycan at Residue 523 of Human Parainfluenza Virus Type
3 Hemagglutinin-Neuraminidase Masks a Second Receptor-Binding Site. Journal of
virology, 2010. 84(6): p. 3094-3100.
201. Xu, G., Three Sialidases: Three Different Products. Journal of the American Chemical
Society, 2011. 133(6): p. 1718-1721.
202. Berry, A.M., R.A. Lock, and J.C. Paton, Cloning and characterization of nanB, a second
Streptococcus pneumoniae neuraminidase gene, and purification of the NanB enzyme
from recombinant Escherichia coli. Journal of Bacteriology, 1996. 178(16): p. 4854-60.
203. Huberman, K., R.W. Peluso, and A. Moscona, Hemagglutinin–Neuraminidase of Human
Parainfluenza 3: Role of the Neuraminidase in the Viral Life Cycle. Virology, 1995.
214(1): p. 294-300.
204. Zhu, X., et al., Crystal structures of two subtype N10 neuraminidase-like proteins from
bat influenza A viruses reveal a diverged putative active site. Proc Natl Acad Sci U S A,
2012. 109(46): p. 18903-18908.
205. de Jong, M.D. and T.T. Hien, Avian influenza A (H5N1). Journal of clinical virology : the
official publication of the Pan American Society for Clinical Virology, 2006. 35(1): p. 2-13.
206. CDC Introduction and Biology of Influenza website at CDC. 2010-11 Influenza Season -
Influenza Prevention and Control Recommendations, 2011.
207. Rota, P.A., Cocirculation of two distinct evolutionary lineages of influenza type B virus
since 1983. Virology (New York, N.Y.), 1990. 175(1): p. 59.
208. Puzelli, S., et al., Changes in the hemagglutinins and neuraminidases of human
influenza B viruses isolated in Italy during the 2001-02, 2002-03, and 2003-04 seasons.
Journal of medical virology, 2004. 74(4): p. 629-40.
138 | P a g e
209. Duan, S., et al., Oseltamivir–Resistant Pandemic H1N1/2009 Influenza Virus Possesses
Lower Transmissibility and Fitness in Ferrets. PLoS Pathog, 2010. 6(7): p. e1001022.
210. Lewis, A.L., Host sialoglycans and bacterial sialidases: a mucosal perspective Mucosal
sialoglycans and bacterial sialidases. Cellular microbiology, 2012. 14(8): p. 1174-1182.
211. Krivan, H.C., Many pulmonary pathogenic bacteria bind specifically to the carbohydrate
sequence GalNAc beta 1-4Gal found in some glycolipids. Proceedings of the National
Academy of Sciences - PNAS, 1988. 85(16): p. 6157.
212. Uchiyama, S., et al., The surface-anchored NanA protein promotes pneumococcal brain
endothelial cell invasion. The Journal of Experimental Medicine, 2009. 206(9): p. 1845-
1852.
213. Kadioglu, A., et al., The role of Streptococcus pneumoniae virulence factors in host
respiratory colonization and disease. Nat Rev Micro, 2008. 6(4): p. 288-301.
214. Soong, G., Bacterial neuraminidase facilitates mucosal infection by participating in
biofilm production. The Journal of clinical investigation, 2006. 116(8): p. 2297.
215. Pettigrew, M.M., Variation in the Presence of Neuraminidase Genes among
Streptococcus pneumoniae Isolates with Identical Sequence Types. Infection and
immunity, 2006. 74(6): p. 3360-3365.
216. Manco, S., Pneumococcal Neuraminidases A and B Both Have Essential Roles during
Infection of the Respiratory Tract and Sepsis. Infection and immunity, 2006. 74(7): p.
4014-4020.
217. Taylor, G., Sialidases: structures, biological significance and therapeutic potential.
Current opinion in structural biology, 1996. 6(6): p. 830-837.
218. Crennell, S.J., et al., Crystal Structure of a Bacterial Sialidase (From Salmonella
typhimurium LT2) Shows the Same Fold as an Influenza Virus Neuraminidase.
Proceedings of the National Academy of Sciences of the United States of America, 1993.
90(21): p. 9852-9856.
139 | P a g e
219. Gut, H., Structural basis for Streptococcus pneumoniae NanA inhibition by influenza
antivirals zanamivir and oseltamivir carboxylate. Journal of molecular biology, 2011.
220. Xu, G., Crystal structure of the NanB sialidase from Streptococcus pneumoniae. Journal
of molecular biology, 2008. 384(2): p. 436.
221. Baker, A.T., et al., Three-dimensional structure of neuraminidase of subtype N9 from an
avian influenza virus. Proteins: Structure, Function, and Bioinformatics, 1987. 2(2): p.
111-117.
222. Moustafa, I., et al., Sialic Acid Recognition by Vibrio cholerae Neuraminidase. Journal of
Biological Chemistry, 2004. 279(39): p. 40819-40826.
223. Peacey, M., et al., Diagnostic Assay Recommended by the World Health Organization
for Swine Origin Influenza A (H1N1) Virus Cross-Reacts with H5N1 Influenza Virus.
Journal of Clinical Microbiology, 2009. 47(11): p. 3789-3790.
224. Duchamp, M.B., et al., Pandemic A(H1N1)2009 influenza virus detection by real time
RT-PCR : is viral quantification useful? Clinical Microbiology and Infection, 2010. 16(4):
p. 317-321.
225. Smith, K.A., et al., High titer growth of human and avian influenza viruses in an
immortalized chick embryo cell line without the need for exogenous proteases. Vaccine,
2008. 26(29): p. 3778-3782.
226. McShane, D., Airway surface pH in subjects with cystic fibrosis. The European
respiratory journal, 2003. 21(1): p. 37-42.
227. Jacobs, S., Studies on the pH of the respiratory tract. The Journal of pediatrics, 1958.
53(3): p. 298-302.
228. Takahashi, T., et al., Duck and Human Pandemic Influenza A Viruses Retain Sialidase
Activity under Low pH Conditions. Journal of Biochemistry, 2001. 130(2): p. 279-283.
140 | P a g e
229. Takahashi, T., et al., The Low-pH Stability Discovered in Neuraminidase of 1918
Pandemic Influenza A Virus Enhances Virus Replication. PLoS ONE, 2010. 5(12): p.
e15556.
230. Takahashi, T., et al., Amino acid determinants conferring stable sialidase activity at low
pH for H5N1 influenza A virus neuraminidase. FEBS Open Bio, 2012. 2: p. 261-266.
231. Air, G.M., The neuraminidase of influenza virus. Proteins, structure, function, and
bioinformatics, 1989. 6(4): p. 341-356.
232. Selimova, L.M., Disulfide bonding in influenza virus proteins as revealed by
polyacrylamide gel electrophoresis. Journal of virology, 1982. 44(2): p. 450.
233. Stigler, J. and M. Rief, Calcium-dependent folding of single calmodulin molecules.
Proceedings of the National Academy of Sciences, 2012.
234. Lytle, B.L., et al., Secondary Structure and Calcium-Induced Folding of the Clostridium
thermocellum Dockerin Domain Determined by NMR Spectroscopy. Archives of
biochemistry and biophysics, 2000. 379(2): p. 237-244.
235. Zhang, L., Calcium-induced Folding and Stabilization of the Pseudomonas aeruginosa
Alkaline Protease. The Journal of biological chemistry, 2012. 287(6): p. 4311-4322.
236. He, G., Clinical pharmacokinetics of the prodrug oseltamivir and its active metabolite Ro
64-0802. Clinical pharmacokinetics, 1999. 37(6): p. 471.
237. Deng, Y.M., A comparison of pyrosequencing and neuraminidase inhibition assays for
the detection of oseltamivir-resistant pandemic influenza A(H1N1) 2009 viruses. Antiviral
Research, 2011. 90(1): p. 87-91.
238. Deyde, V.M., et al., Detection of Molecular Markers of Drug Resistance in 2009
Pandemic Influenza A (H1N1) Viruses by Pyrosequencing. Antimicrobial agents and
chemotherapy, 2010. 54(3): p. 1102-1110.
141 | P a g e
239. Guo, L., et al., Rapid identification of oseltamivir-resistant influenza A(H1N1) viruses with
H274Y mutation by RT-PCR/restriction fragment length polymorphism assay. Antiviral
Research, 2009. 82(1): p. 29-33.
240. Tong, S.Y., et al., Rapid detection of the H275Y oseltamivir resistance mutation in
influenza A/H1N1 2009 by single base pair RT-PCR and high-resolution melting. PLoS
ONE, 2011. 6(6): p. e21446.
241. Wu, Z.L., et al., Active 1918 pandemic flu viral neuraminidase has distinct N-glycan
profile and is resistant to trypsin digestion. Biochemical and Biophysical Research
Communications, 2009. 379(3): p. 749-753.
242. Bonner, A.B., et al., Impact of the Rapid Diagnosis of Influenza on Physician Decision-
Making and Patient Management in the Pediatric Emergency Department: Results of a
Randomized, Prospective, Controlled Trial. Pediatrics, 2003. 112(2): p. 363-367.
243. Noyola, D.E. and G.J. Demmler, Effect of rapid diagnosis on management of influenza A
infections. The Pediatric Infectious Disease Journal, 2000. 19(4): p. 303-307.
244. Richards, S. Evaluation of Rapid Influenza Diagnostic Tests for Influenza A (H3N2)v
Virus and Updated Case Count. Morbidity and Mortality Weekly Report, 2012. 61, 619-
621.
245. Koh, X., Glycan receptor binding determinants of Influenza A virus hemagglutinin, 2010,
Massachusetts Institute of Technology.
246. Stanley, P., H. Schachter, and N. Taniguchi, Chapter 8: N-Glycans, in Essentials of
Glycobiology, A. Varki, R. Cummings, and J. Esko, Editors. 2009, Cold Spring Harbor
Laboratory Press: NY.
247. Watanabe, T., S. Watanabe, and Y. Kawaoka, Cellular networks involved in the influenza
virus life cycle. Cell Host Microbe, 2010. 7(6): p. 427-39.
248. Feinberg, H., et al., Structural basis for selective recognition of oligosaccharides by DC-
SIGN and DC-SIGNR. Science, 2001. 294(5549): p. 2163-6.
142 | P a g e
249. Guo, Y., Structural basis for distinct ligand-binding and targeting properties of the
receptors DC-SIGN and DC-SIGNR. Nature structural & molecular biology, 2004. 11(7):
p. 591-598.
250. Powlesland, A.S., et al., Widely divergent biochemical properties of the complete set of
mouse DC-SIGN-related proteins. The Journal of biological chemistry, 2006. 281(29): p.
20440-9.
251. Caminschi, I., et al., Functional comparison of mouse CIRE/mouse DC-SIGN and human
DC-SIGN. International immunology, 2006. 18(5): p. 741-53.
252. CFG, Consortium for Functional Glycomics, in CFG Molecule pages - Glycan Binding
Protein - SIGNR52010, Functional Glycomics Gateway.
253. Takahashi, K., et al., The mannose-binding lectin: a prototypic pattern recognition
molecule. Current Opinion in Immunology, 2006. 18(1): p. 16-23.
254. Ng, W.C., et al., Soluble host defense lectins in innate immunity to influenza virus.
Journal of biomedicine & biotechnology, 2012. 2012: p. 732191.
255. Yang, M.L., et al., Galectin-1 binds to influenza virus and ameliorates influenza virus
pathogenesis. Journal of virology, 2011. 85(19): p. 10010-20.
256. Chernyy, E.S., et al., Galectins promote the interaction of influenza virus with its target
cell. Biochemistry. Biokhimiia, 2011. 76(8): p. 958-67.
257. Wang, Y.F., et al., Characterization of glycan binding specificities of influenza B viruses
with correlation with hemagglutinin genotypes and clinical features. Journal of medical
virology, 2012. 84(4): p. 679-85.
258. Cheng, X., et al., Surface glycoproteins of influenza A H3N2 virus modulate virus
replication in the respiratory tract of ferrets. Virology, 2012. 432(1): p. 91-8.
259. Ohuchi, M., et al., Regulation of receptor binding affinity of influenza virus hemagglutinin
by its carbohydrate moiety. J Virol, 1997. 71(11): p. 8377 - 8384.
143 | P a g e
260. Tate, M.D., et al., Glycosylation of the hemagglutinin modulates the sensitivity of H3N2
influenza viruses to innate proteins in airway secretions and virulence in mice. Virology,
2011. 413(1): p. 84-92.
261. Stroop, C.J.M., et al., Characterization of the carbohydrate chains of the secreted form of
the human epidermal growth factor receptor. Glycobiology, 2000. 10(9): p. 901-917.
262. Vasta, G.R., Roles of galectins in infection. Nature reviews. Microbiology, 2009. 7(6): p.
424-38.
263. CDC. Enterohemorrhagic Escherichia coli. National Center for Zoonotic, Vector-Borne,
and Enteric Diseases 2005 July 21, 2010 [cited 2011; Available from:
http://www.cdc.gov/nczved/divisions/dfbmd/diseases/enterohemorrhagic_ecoli/technical.
html.
264. Bitzan, M., Treatment options for HUS secondary to Escherichia coli O157:H7. Kidney
Int, 2009. 75(112s): p. S62-S66.
265. Whyte, D.A. and R.N. Fine, Chronic Kidney Disease in Children. Pediatrics in Review,
2008. 29(10): p. 335-341.
266. Okuda, T., et al., Targeted Disruption of Gb3/CD77 Synthase Gene Resulted in the
Complete Deletion of Globo-series Glycosphingolipids and Loss of Sensitivity to
Verotoxins. Journal of Biological Chemistry, 2006. 281(15): p. 10230-10235.
267. Shimizu, T., et al., Receptor affinity, stability and binding mode of Shiga toxins are
determinants of toxicity. Microbial Pathogenesis, 2007. 43(2-3): p. 88-95.
268. Shin, I.-S., et al., Globotriaosylceramide (Gb3) content in HeLa cells is correlated to
Shiga toxin-induced cytotoxicity and Gb3 synthase expression BMB reports, 2009.
42(5): p. 310-314.
269. Endo, Y., et al., Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of
Shiga toxin on eukaryotic ribosomes. European Journal of Biochemistry, 1988. 171(1-2):
p. 45-50.
144 | P a g e
270. Obrig, T.G., T.P. Moran, and J.E. Brown, The mode of action of Shiga toxin on peptide
elongation of eukaryotic protein synthesis. Biochem. J., 1987. 244(2): p. 287-294.
271. Tesh, V.L. and A.D. O'Brien, The pathogenic mechanisms of Shiga toxin and the Shiga-
like toxins. Molecular Microbiology, 1991. 5(8): p. 1817-1822.
272. Kozlov, Y.V., et al., The primary structure of the operons coding for Shigella dysenteriae
toxin and temperate phage H30 shiga-like toxin. Gene, 1988. 67(2): p. 213-221.
273. Fuller, C.A., et al., Shiga toxin subtypes display dramatic differences in potency. Infect.
Immun., 2011: p. IAI.01182-10.
274. Rutjes, N.W.P., et al., Differential tissue targeting and pathogenesis of verotoxins 1 and
2 in the mouse animal model. Kidney Int, 2002. 62(3): p. 832-845.
275. Tesh, V.L., et al., Comparison of the relative toxicities of Shiga-like toxins type I and type
II for mice. Infect. Immun., 1993. 61(8): p. 3392-3402.
276. Head, S.C., M.A. Karmali, and C.A. Lingwood, Preparation of VT1 and VT2 hybrid toxins
from their purified dissociated subunits. Evidence for B subunit modulation of a subunit
function. Journal of Biological Chemistry, 1991. 266(6): p. 3617-3621.
277. Flagler, M.J., et al., Comparison of Binding Platforms Yields Insights into Receptor
Binding Differences between Shiga Toxins 1 and 2. Biochemistry, 2010. 49(8): p. 1649-
1657.
278. Weinstein, D.L., et al., In vivo formation of hybrid toxins comprising Shiga toxin and the
Shiga-like toxins and role of the B subunit in localization and cytotoxic activity. Infect.
Immun., 1989. 57(12): p. 3743-3750.
279. Lingwood, C.A., Role of verotoxin receptors in pathogenesis. Trends in microbiology,
1996. 4(4): p. 147-153.
280. Armstrong, G.D., et al., Human Serum Amyloid P Component Protects against
Escherichia coli O157:H7 Shiga Toxin 2 In Vivo: Therapeutic Implications for Hemolytic-
Uremic Syndrome. Journal of Infectious Diseases, 2006. 193(8): p. 1120-1124.
145 | P a g e
281. Ling, H., et al., Structure of the Shiga-like Toxin I B-Pentamer Complexed with an
Analogue of Its Receptor Gb3†,‡. Biochemistry, 1998. 37(7): p. 1777-1788.
282. Cummings, M.D., et al., Modeling the Carbohydrate-Binding Specificity of Pig Edema
Toxin†. Biochemistry, 1998. 37(7): p. 1789-1799.
283. Ling, H., et al., A mutant Shiga-like toxin IIe bound to its receptor Gb3: structure of a
group II Shiga-like toxin with altered binding specificity. Structure, 2000. 8(3): p. 253-264.
284. Nakajima, H., et al., Kinetic Analysis of Binding between Shiga Toxin and Receptor
Glycolipid Gb3Cer by Surface Plasmon Resonance. Journal of Biological Chemistry,
2001. 276(46): p. 42915-42922.
285. Kale, R.R., et al., Differentiation between Structurally Homologous Shiga 1 and Shiga 2
Toxins by Using Synthetic Glycoconjugates. Angewandte Chemie International Edition,
2008. 47(7): p. 1265-1268.
286. Miura, Y., et al., Self-assembled monolayers of globotriaosylceramide (Gb3) mimics:
surface-specific affinity with shiga toxins. Analytical Biochemistry, 2002. 310(1): p. 27-35.
287. Flagler, M.J., Determination of the Molecular Basis for the Difference in Potency between
Shiga Toxins 1 and 2, in Molecular Genetics, Biochemistry, & Microbiology2010,
University of Cincinnati: Cincinnati. p. 242.
288. Kulkarni, A.A., A.A. Weiss, and S.S. Iyer, Glycan-based high-affinity ligands for toxins
and pathogen receptors. Medicinal Research Reviews, 2010. 30(2): p. 327-393.
289. Waddell, T., et al., Globotriosyl ceramide is specifically recognized by the escherichia
coli verocytotoxin 2. Biochemical and Biophysical Research Communications, 1988.
152(2): p. 674-679.
290. DeGrandis, S., et al., Globotetraosylceramide is recognized by the pig edema disease
toxin. Journal of Biological Chemistry, 1989. 264(21): p. 12520-12525.
146 | P a g e
291. Zumbrun, S.D., et al., Human Intestinal Tissue and Cultured Colonic Cells Contain
Globotriaosylceramide Synthase mRNA and the Alternate Shiga Toxin Receptor,
Globotetraosylceramide. Infect. Immun., 2010: p. IAI.00620-10.
292. Rinaldi, S., et al., Analysis of lectin binding to glycolipid complexes using combinatorial
glycoarrays. Glycobiology, 2009. 19(7): p. 789-796.
293. Jacobson, K., O.G. Mouritsen, and R.G.W. Anderson, Lipid rafts: at a crossroad between
cell biology and physics. Nat Cell Biol, 2007. 9(1): p. 7-14.
294. Risselada, H.J. and S.J. Marrink, The molecular face of lipid rafts in model membranes.
Proceedings of the National Academy of Sciences, 2008. 105(45): p. 17367-17372.
295. Wen, S.X., et al., Genetic toxoids of Shiga toxin types 1 and 2 protect mice against
homologous but not heterologous toxin challenge. Vaccine, 2006. 24(8): p. 1142-1148.
296. Conrady, D.G., et al., Molecular Basis of Differential B-Pentamer Stability of Shiga
Toxins 1 and 2. PLoS ONE, 2010. 5(12): p. e15153.
297. McGannon, C.M., C.A. Fuller, and A.A. Weiss, Different Classes of Antibiotics
Differentially Influence Shiga Toxin Production. Antimicrob. Agents Chemother., 2010.
54(9): p. 3790-3798.
298. St. Hilaire, P.M., M.K. Boyd, and E.J. Toone, Interaction of the Shiga-like Toxin Type 1
B-Subunit with Its Carbohydrate Receptor. Biochemistry, 1994. 33(48): p. 14452-14463.
299. Kitova, E.N., et al., Affinities of Shiga toxins 1 and 2 for univalent and oligovalent Pk-
trisaccharide analogs measured by electrospray ionization mass spectrometry.
Glycobiology, 2007. 17(10): p. 1127-1137.
300. Binnington, B., et al., Effect of Globotriaosyl Ceramide Fatty Acid α-Hydroxylation on the
Binding by Verotoxin 1 and Verotoxin 2. Neurochemical Research, 2002. 27(7): p. 807-
813.
301. Lingwood, C.A., et al., Globotriaosyl ceramide receptor function – Where membrane
structure and pathology intersect. FEBS letters, 2010. 584(9): p. 1879-1886.
147 | P a g e
302. Boyd, B., et al., Lipid modulation of glycolipid receptor function. European Journal of
Biochemistry, 1994. 223(3): p. 873-878.
303. Donohue-Rolfe, A., D.W.K. Acheson, and G.T. Keusch, Shiga Toxin: Purification,
Structure, and Function. Review of Infectious Diseases, 1991. 13(Supplement 4): p.
S293-S297.
304. Millen, S.H., et al., Identification and Characterization of the Carbohydrate Ligands
Recognized by Pertussis Toxin via a Glycan Microarray and Surface Plasmon
Resonance. Biochemistry, 2010. 49(28): p. 5954-5967.
305. Lewallen, D.M., D. Siler, and S.S. Iyer, Factors Affecting Protein–Glycan Specificity:
Effect of Spacers and Incubation Time. ChemBioChem, 2009. 10(9): p. 1486-1489.
306. Heinrich, L., et al., Comparison of the results obtained by ELISA and surface plasmon
resonance for the determination of antibody affinity. Journal of Immunological Methods,
2010. 352(1-2): p. 13-22.
307. Norton, W.T., et al., The lipid composition of isolated brain cells and axons. Journal of
Neuroscience Research, 1975. 1(1): p. 57-75.
308. Ogawa-Goto, K. and T. Abe, Gangliosides and Glycosphingolipids of Peripheral Nervous
System Myelins—a Minireview. Neurochemical Research, 1998. 23(3): p. 305-310.
309. Basta, M., M. Karmali, and C. Lingwood, Sensitive receptor-specified enzyme-linked
immunosorbent assay for Escherichia coli verocytotoxin. J. Clin. Microbiol., 1989. 27(7):
p. 1617-1622.
310. Takenaga, M., et al., Effect of a soluble pseudo-receptor on verotoxin 2-induced toxicity.
Journal of Infection and Chemotherapy, 2000. 6(1): p. 21-25.
311. Boulanger, J., et al., Neutralization receptor-based immunoassay for detection of
neutralizing antibodies to Escherichia coli verocytotoxin 1. J. Clin. Microbiol., 1990.
28(12): p. 2830-2833.
148 | P a g e
312. Kim, Y.-H., et al., Distribution of Verocytotoxin - Producing Escherichia coli and
Development of Receptor Binding Enzyme - Linked Immunosorbent Assay 2. Korean
Journal of Veterenerian Public Heatlh, 1996. 20(1): p. 55.
313. Pellizzari, A., H. Pang, and C.A. Lingwood, Binding of verocytotoxin 1 to its receptor is
influenced by differences in receptor fatty acid content. Biochemistry, 1992. 31(5): p.
1363-1370.
314. Khan, F., F. Proulx, and C.A. Lingwood, Detergent-resistant globotriaosyl ceramide may
define verotoxin/glomeruli-restricted hemolytic uremic syndrome pathology. Kidney Int,
2009. 75(11): p. 1209-1216.
315. Yahi, N., A. Aulas, and J. Fantini, How Cholesterol Constrains Glycolipid Conformation
for Optimal Recognition of Alzheimer's β Amyloid Peptide (Aβ1-40). PLoS
ONE, 2010. 5(2): p. e9079.
316. Lingwood, D., et al., Cholesterol modulates glycolipid conformation and receptor activity.
Nat Chem Biol, 2011. 7(5): p. 260-262.
317. Mahfoud, R., et al., A Major Fraction of Glycosphingolipids in Model and Cellular
Cholesterol-containing Membranes Is Undetectable by Their Binding Proteins. Journal of
Biological Chemistry, 2010. 285(46): p. 36049-36059.
318. Hammache, D., et al., Specific Interaction of HIV-1 and HIV-2 Surface Envelope
Glycoproteins with Monolayers of Galactosylceramide and Ganglioside GM3. Journal of
Biological Chemistry, 1998. 273(14): p. 7967-7971.
319. Scheiring, J., S. Andreoli, and L. Zimmerhackl, Treatment and outcome of Shiga-toxin-
associated hemolytic uremic syndrome (HUS). Pediatric Nephrology, 2008. 23(10): p.
1749-1760.
320. Katagiri, Y.U., et al., Activation of Src Family Kinase Yes Induced by Shiga Toxin Binding
to Globotriaosyl Ceramide (Gb3/CD77) in Low Density, Detergent-insoluble
Microdomains. Journal of Biological Chemistry, 1999. 274(49): p. 35278-35282.
149 | P a g e
321. Liu, F., J. Huang, and J.E. Sadler, Shiga toxin (Stx)1B and Stx2B induce von Willebrand
factor secretion from human umbilical vein endothelial cells through different signaling
pathways. Blood, 2011. 118(12): p. 3392-3398.
322. Huang, J., et al., Shiga toxin B subunits induce VWF secretion by human endothelial
cells and thrombotic microangiopathy in ADAMTS13-deficient mice. Blood, 2010.
116(18): p. 3653-3659.
323. Palermo, M.S., R.A. Exeni, and G.C. Fernández, Hemolytic uremic syndrome:
pathogenesis and update of interventions. Expert review of anti-infective therapy, 2009.
7(6): p. 697-707.
324. Ngaosuwankul, N., et al., Influenza A viral loads in respiratory samples collected from
patients infected with pandemic H1N1, seasonal H1N1 and H3N2 viruses, 2010, BioMed
Central.
325. WHO, Collecting, preserving and shipping specimens for the diagnosis of avian influenza
A(H5N1) virus infection. Guide for field operations, G.A.a.R. (GAR), Editor 2006, GAR
publications
326. Hayre, J., et al., Optimization of a direct spectrophotometric method to investigate the
kinetics and inhibition of sialidases. BMC Biochemistry, 2012. 13(1): p. 19.
327. Hurt, A.C., et al., Zanamivir-resistant influenza viruses with a novel neuraminidase
mutation. J Virol, 2009. 83(20): p. 10366-73.
328. Calfee, D.P., et al., Safety and Efficacy of Intravenous Zanamivir in Preventing
Experimental Human Influenza A Virus Infection. Antimicrobial Agents and
Chemotherapy, 1999. 43(7): p. 1616-1620.
329. Cass, L.M.R., et al., Pharmacoscintigraphic evaluation of lung deposition of inhaled
zanamivir in healthy volunteers. Clinical Pharmacokinetics, 1999. 36(SUPPL. 1): p. 21-
31.
150 | P a g e
330. CDC Seasonal Influenza (Flu) - H3N2v and You - website at CDC. Influenza A (H3N2)
Variant Virus Outbreaks - H3N2v and You, 2012.
331. Gubareva, L.V., Detection of influenza virus resistance to neuraminidase inhibitors by an
enzyme inhibition assay. Antiviral Research, 2002. 53(1): p. 47-61.
332. Suzuki, T., et al., Evolutional analysis of human influenza A virus N2 neuraminidase
genes based on the transition of the low-pH stability of sialidase activity. FEBS Letters,
2004. 557(1-3): p. 228-232.
333. Takahashi, T., et al., A molecular mechanism for the low-pH stability of sialidase activity
of influenza A virus N2 neuraminidases. FEBS Letters, 2003. 543(1–3): p. 71-75.
334. Xu, G., et al., Three Streptococcus pneumoniae sialidases: three different products. J
Am Chem Soc, 2011. 133(6): p. 1718-21.
335. Telford, J.C., et al., The Aspergillus fumigatus sialidase is a 3-deoxy-D-glycero-D-
galacto-2-nonulosonic acid hydrolase (KDNase): structural and mechanistic insights. J
Biol Chem, 2011. 286(12): p. 10783-92.
336. Chan, J., et al., Transition state analysis of Vibrio cholerae sialidase-catalyzed
hydrolyses of natural substrate analogues. J Am Chem Soc, 2012. 134(8): p. 3748-57.
337. Lentz, M.R., Site-directed mutation of the active site of influenza neuraminidase and
implications for the catalytic mechanism. Biochemistry (Easton), 1987. 26(17): p. 5351-
5358.
338. Burmeister, W.P., et al., Influenza B virus neuraminidase can synthesize its own
inhibitor. Structure (London, England : 1993), 1993. 1(1): p. 19-26.
339. Chong, A.K.J., et al., Evidence for a sialosyl cation transition-state complex in the
reaction of sialidase from influenza virus. European Journal of Biochemistry, 1992.
207(1): p. 335-343.
340. Vavricka, C.J., et al., Influenza neuraminidase operates via a nucleophilic mechanism
and can be targeted by covalent inhibitors. Nat Commun, 2013. 4: p. 1491.
151 | P a g e
341. Kim, J.-H., et al., Mechanism-Based Covalent Neuraminidase Inhibitors with Broad
Spectrum Influenza Antiviral Activity. Science, 2013.
342. Watts, A.G. and S.G. Withers, The synthesis of some mechanistic probes for sialic acid
processing enzymes and the labeling of a sialidase fromTrypanosoma rangeli. Canadian
Journal of Chemistry, 2004. 82(11): p. 1581-1588.
343. Harris, T.K. and G.J. Turner, Structural Basis of Perturbed pKa Values of Catalytic
Groups in Enzyme Active Sites. IUBMB Life, 2002. 53(2): p. 85-98.
344. Nielsen, J.E. and J.A. McCammon, Calculating pKa values in enzyme active sites.
Protein Science, 2003. 12(9): p. 1894-1901.
345. McIntosh, L.P., et al., The pKa of the General Acid/Base Carboxyl Group of a
Glycosidase Cycles during Catalysis: A 13C-NMR Study of Bacillus circulans
Xylanase†. Biochemistry, 1996. 35(31): p. 9958-9966.
346. Highbarger, L.A., J.A. Gerlt, and G.L. Kenyon, Mechanism of the Reaction Catalyzed by
Acetoacetate Decarboxylase. Importance of Lysine 116 in Determining the pKa of
Active-Site Lysine 115†. Biochemistry, 1996. 35(1): p. 41-46.
347. de Lederkremer, R.M. and R. Agusti, Chapter 7 Glycobiology of Trypanosoma cruzi, in
Advances in Carbohydrate Chemistry and Biochemistry, H. Derek, Editor. 2009,
Academic Press. p. 311-366.
348. Bose, S., et al., Fusion activation by a headless parainfluenza virus 5 hemagglutinin-
neuraminidase stalk suggests a modular mechanism for triggering. Proceedings of the
National Academy of Sciences, 2012. 109(39): p. E2625–E2634.
349. Rydbeck, R., A. Löve, and C. Örvell, Parainfluenza Virus Type 2 Haemagglutinin-
Neuraminidase Glycoprotein Characterized with Monoclonal Antibodies. Journal of
General Virology, 1988. 69(4): p. 931-935.
350. Cámara, M., et al., A neuraminidase from Streptococcus pneumoniae has the features of
a surface protein. Infection and Immunity, 1994. 62(9): p. 3688-3695.
152 | P a g e
351. Colacino, J.M., et al., A Single Sequence Change Destabilizes the Influenza Virus
Neuraminidase Tetramer. Virology, 1997. 236(1): p. 66-75.
153 | P a g e
APPENDIX
1. Effect of pH on Neuraminidase activity
Previous studies indicate that low-pH activity of influenza A virus neuraminidase is critical for viral replication in birds; while it is less essential for replication of human influenza strains.
[228, 332]. More recently, it has been reported that influenza strains with NA activity at low pH might have a higher replication rate and virulence [14]. Our findings show that human N1 and
Yamagata strains display high activity at pH 5 while human N2 and Victoria strains display no activity at pH 5. In addition, we found that Streptococcus pneumonia NA activity has a broad range of pH 5-9, while parainfluenza NA activity is limited to pH 4-5.
Previously, sequence analysis of several influenza neuraminidases show that there is not a consensus amino acid domain responsible for low-pH influenza NA enzymatic activity [229,
333], however, single residues located at the interphase of monomers were identified in N1
[229] and N2 tetramers [333] to play a role in the influenza NA activity at low pH.
We compared the 3D structure of NA in order to gain insights into the effect of pH on neuraminidase activity. We observed that all NAs tested in this study have the usual catalytic residues common to all neuraminidases. These contain three arginines that interact with the carboxylate group of NeuAc, a nucleophilic tyrosine and a glutamic acid. These shared catalytic residues are also found in many sialidases, which cleave NeuAc producing different NeuAc analogs (e.g. streptococcal NanB and NanC; [334]). In addition, these common residues are also found in neuraminidases with very different optimal pH activities such as neuraminidases from Aspergillus fumigatus (optimal pH 3.5; [335]), streptococcal NanB (optimal pH 4.5), Vibrio cholerae (optimal pH 5.5 [336]) and Trypanosoma cruzi (optimal pH 6.5-7.5 [220]).
We observed that the NeuAc binding site in each of the NAs considered in this study contains a negatively charged pocket and positively charged region as previously reported [124]
(Figure 6.1). In the N1 structure, Glu119, Ser179, Ser181, Glu227, and Glu277 form the
154 | P a g e
negatively charged pocket and Arg118 and Arg371 form the positively charged region (Figure
6.1A). In the N2 structure, Glu119, Ser179, Ser180, Glu227, and Glu277 form the negatively charged pocket and Arg118 and Arg371 form the positively charged region (Figure 6.1B). In parainfluenza HN, Glu276, Glu286, Tyr319, and Tyr337 form the negatively charged pocket and
Arg192, Arg502, Arg424 form the positively charged region. In streptococcus NanA, Asn357,
Asp402, and Asn419 form the negatively charged pocket and Arg332, Arg648, Arg706 form the positively charge region.
A. Influenza N1 B. Influenza N2
C. Parainfluenza HN D. S. pneumoniae NanA
Figure 6.1. Electrostatic potential map of neuraminidase. A. Influenza N1 neuraminidase (PDB 2HU4) in complex with oseltamivir (shown in cyan). B. Influenza N2 neuraminidase (PDB 2BAT) in complex with NeuAc (shown in cyan). C. Parainfluenza HN (PDB 1V3C) in complex with NeuAc (shown in cyan). D. S. pneumoniae NanA (PDB 2YA5) in complex with NeuAc (shown in cyan). The protein surface charge is shown, where red is a negatively charged area, blue is a positively charged area and white is a neutral hydrophobic area.
At the time of my defense, the mechanism of action of neuraminidase enzyme had not been published. Decades ago, it was originally suggested that the enzymatic mechanism of influenza virus sialidase included the formation of an oxocarbocation ion intermediate, a sialosyl cation [141, 231, 337-339]. Since this writing, strong evidence was shown to support that the
155 | P a g e
catalytic mechanism of influenza NA for cleaving NeuAc includes the formation of a sialoside- enzyme covalently bound intermediate, which uses the tyrosine residue as the catalytic nucleophile (Figure 6.2). Those experiments were done at physiological pH only [340, 341].
Equivalent catalytic mechanisms, involving formation of tyrosine bound intermediate, have been reported at neutral pH for Trypanosoma trans-sialidase, an enzyme responsible for transferring
NeuAc from host to parasite, [342] and Vibrio neuraminidase at pH 5.5 [336]. However, Vibrio neuraminidase is suggested to additionally form a transient oxocarbenium ion at pH 5.5 [336].
These reports demonstrate that the mechanism of cleaving NeuAc at neutral or low pH is equivalent for members of the neuramidase family. However, more research is needed to understand the mechanism of influenza NA activity at low pH.
In influenza NA substrate-binding site, only a small number of different side-chain groups function as catalysts. In order to catalyze a reaction, these side-chain groups need to be in a functional protonation state. Regularly, ionization of side-chain groups depends on the intrinsic pKa of the group and the microenvironment in the protein which surrounds the side-chain group.
These two factors perturb intrinsic pKa of the side-chain group to its functional pKa [343].
In influenza NA, the residues located in the negatively charged pocket have an intrinsic pKa of
~4.3; while the residues located in the positively charged pocket have an intrinsic pKa of ~12.5, while this suggests that the side chains of these residues will be in a functional ionization state in our assay (pH 5-9), usually, pKas of side-chain groups of catalytic residues are considerably perturbed (more than 2 pKa units) in the active site [343]. Ionization of a catalytic group mainly depends on (1) the distance between this group and other near charged groups and (2) the dielectric constant of the medium between them [343, 344]. The effect of proximal negatively or positively charged groups on the functional pKas are well established on other enzymatic mechanisms [345, 346]. However, no information is available on the functional pKas of catalytic groups in the influenza NA active site.
156 | P a g e
Figure 6.2. Scheme of the mechanisms of influenza NA. Glu227 activates a catalytic nucleophile residue, Tyr377 (Step 1). Upon binding, there is a distortion of the NeuAc sugar ring to a boat conformation (Step 2). Subsequently, a covalently bonded intermediate is formed in chair conformation (Step 3). The negatively charged binding pocket stabilizes the charged intermediate. This intermediate is hydrolytically released as NeuAc (Step 5). Figure obtained from de Lederkremer et al. (2009) and modified based on Vavricka et al. (2013) and Kim et al. (2013).
From the recently published mechanism of influenza NA, Tyr406, Glu277 and Asp151 are identified as the catalytic residues [341], where Glu and Asp need to be in functional ionization states (Figure 6.2; [340, 341, 347]). Several charged residues are located near these three catalytic residues, such as Glu119, Glu276, Arg118, Arg152, Arg292, and Arg371, all of which might be perturbing the intrinsic pKa of catalytic residues. If the resulting perturbed pKas of
Glu277 (intrinsic pKa of 3.9) or Asp151 (intrinsic pKa of 4.0) are increased due to interaction with surrounding residues, it would affect the range of pH activity of influenza NA. Since the distance between a catalytic group and other near charged group perturbs intrinsic pKas, we measured the distance between each of site-chain groups of surrounding residues and the
157 | P a g e
functional groups in catalytic residues for influenza NAs. We compared these distances between
N1 (PBD 2HU4), N2 (PBD 2BAT) and type B NA (PDB 1A4G), and observed no significant difference in each distance. This observation does not explain the low-pH activity of influenza
N1 and not N2. We believed that the perturbed pKa of Glu277 and Asp151 should be determined in order to gain some understanding in the mechanism of low pH activity of influenza
NA.
Evidence has suggested that low pH activity of neuraminidase is mediated by stabilization of the NA multimer [229, 333]. Influenza NA tetramer is formed by four identical not-covalently bound monomers each one with a NeuAc binding site. The stalk of each monomer is attached to the viral membrane [124, 134]. Parainfluenza hemagglutinin-neuraminidase (HN) is reported to form a dimer of dimers, each of which is linked by disulfide bonds [348, 349]. HN dimers are also attached to the viral membrane through a stalk region. Streptococcus NanA is suggested to form a dimer attached to the surface of the bacteria. Studies with NA of H1N1 influenza strain demonstrated that deletion of Thr435 or deletion of Gly455 or a mutation of Arg430 to Gln stabilizes enzymatic activity at low pH [229]. In addition, studies with NA of H3N2 influenza strain indicated that substitutions of Arg344 to Lys and Phe344 to Leu reduced the stability of
NA activity at low pH [333]. These residues on N1 and N2 are located in the subunit-interface and substitutions of these residues could lead to a conformational change in the overall tetramer form of the influenza NA structure. Taken together, these results strongly suggest that the influenza NA activity at low pH depends on the stability of NA the structure of the native NA tetramer. Parainfluenza HN dimers are covalently bound to each other providing strong stability of HN structure, while streptococcus NanA is covalently anchored to the bacterial cell wall [350], which stabilizes the structure of the enzyme. Having a strong stable structure, HN shows activity at pH 4-5, and NanA shows activity a pH range of 6-9.
158 | P a g e
The stability of influenza NA tetramer is less than that of HN and NanA. Therefore, pH variations could potentially affect the influenza NA oligomeric conformation. Previous studies showed that monomeric and dimeric forms of influenza NA have reduced activity and low affinity for NeuAc [241]. Moreover, single point mutations in the subunit interface destabilize active influenza NA tetramer into non-active monomers suggesting that the weak interaction between free monomers affects enzymatic activity [351]. Consequently, lower order oligomers of influenza NA present at low pH ranges could result in reduced enzymatic activity. In this study, only N1 and Yamagata strains show activity at pH 5. To test this hypothesis, future studies could confirm that these strains have a strong interaction of NA monomers, which is needed to retain enzymatic activity at low pH.
159 | P a g e
2. Sequence alignments of NA genes of H1N1pdm strains
Sequence alignment were perform using that basic local alignment search tool (BLAST) from NBCI; single-letter amino acid code is used; X is used for unknown amino acid residues.
09-S 1 MNPNQKIITI GSVCMTIGMA NLILQIGNII SIWISHSIQL GNQNQIETCN QSVITYENNT 60
09-R 1 ...... 60
09-S 61 WVNQTYVNIS NTNFAAGQSV VSVKLAGNSS LCPVSGWAIY SKDNSVRIGS KGDVFVIREP 120
09-R 61 ...... I...... 120
09-S 121 FISCSPLECR TFFLTQGALL NDKHSNGTIK DRSPYRTLMS CPIGEVPSPY NSRFESVAWS 180
09-R 121 ...... 180
09-S 181 ASACHDGINW LTIGISGPDN GAVAVLKYNG IITDTIKSWR NNILRTQESE CACVNGSCFT 240
09-R 181 ...... 240
09-S 241 VMTDGPSNGQ ASYKIFRIEK GKIVKSVEMN APNYHYEECS CYPDSSEITC VCRDNWHGSN 300
09-R 241 ...... D...... X. ....Y...... 300
09-S 301 RPWVSFNQNL EYQIGYICSG IFGDNPRPND KTGSCGPVSS NGANGVKGFS FKYGNGVWIG 360
09-R 301 ...... 360
09-S 361 RTKSISSRNG FEMIWDPNGW TGTDNNFSIK QDIVGINEWS GYSGSFVQHP ELTGLDCIRP 420
09-R 361 ...... 420
160 | P a g e
09-S 421 CFWVELIRGR PKENTIWTSG SSISFCGVNS DTVGWSWPDG AELPFTIDK 469
09-R 421 ...... 469
3. Sequence alignments of NA genes of human H1N1 strains
07-R 1 MNPNQKIITI GSISIAIGII SLMLQIGNII SIWASHSIQT GSQNNTGICN QRIITYENST 60
08-R 1 ...... 60
07-R 61 WVNHTYVNIN NTNVVAGEDK TSVTLAGNSS LCSISGWAIY TKDNSIRIGS KGDVFVIREP 120
08-R 61 ...... 120
07-R 121 FISCSHLECR TFFLTQGALL NDKHSNGTVK DRSPYRALMS CPLGEAPSPY NSKFESVAWS 180
08-R 121 ...... 180
07-R 181 ASACHDGMGW LTIGISGPDN GAVAVLKYNG IITGTIKSWK KQILRTQESE CVCMNGSCFT 240
08-R 181 ...... 240
07-R 241 IMTDGPSNKA ASYKIFKIEK GKVTKSIELN APNFHYEECS CYPDTGIVMC VCRDNWHGSN 300
08-R 241 ...... I...... Y...... 300
07-R 301 RPWVSFNQNL DYQIGYICSG VFGDNPRPED GEGSCNPVTV DGANGVKGFS YKYDNGVWIG 360
08-R 301 ...... G...... 360
07-R 361 RTKSNRLRKG FEMIWDPNGW TNTDSDFSVK QDVVAITDWS GYSGSFVQHP ELTGLDCIRP 420
08-R 361 ...... 420
07-R 421 CFWVELVRGL PRENTTIWTS GSSISFCGVN SDTANWSWPD GAELPFTIDK 470
08-R 421 ...... 470
161 | P a g e
4. Sequence alignments of NA genes of H3N2 strains
07-S 1 MNPNQKIITI GSVSLTISTI CFFMQIAILI TTVTLHFKQY EFNSPPNNQV MLCEPTIIER 60
07-R 1 ...... A...... T...... 60
07-S 61 NITEIVYLTN TTIEKEICPK LAEYRNWSKP QCDITGFAPF SKDNSIRLSA GGDIWVTREP 120
07-R 61 ...... V. 120
07-S 121 YVSCDPDKCY QFALGQGTTL NNVHSNDTVR DRTPYRTLLM NELGVPFHLG TKQVCIAWSS 180
07-R 121 ...... 180
07-S 181 SSCHDGKAWL HVCITGDDKN ATASFIYNGR LVDSIVSWSK EILRTQESEC VCINGTCTVV 240
07-R 181 ...... 240
07-S 241 MTDGSASGKA DTKILFIEEG KIVHTSTLSG SAQHVEECSC YPRYPGVRCV CRDNWKGSNR 300
07-R 241 ...... 300
07-S 301 PIVDINIKDH STVSSYVCSG LVGDTPRKND SSSSSHCLDP NNEEGGHGVK GWAFDDGNDV 360
07-R 301 ...... I...... 360
07-S 361 WMGRTISEKS RLGYETFKVI EGWSNPKSKL QINRQVIVDR GNRSGYSGIF SVEGKSCINR 420
07-R 361 ...... 420
07-S 421 CFYVELIRGR KEETEVLWTS NSIVVFCGTS GTYGTGSWPD GADINLMPI 469
07-R 421 ...... K...... 469
162 | P a g e
5. Sequence alignments of NA genes of influenza type B strains
Vict/06-S 1 MLPSTIQTLT LFLTSGGVLL SLYVSASLSY LLYSDILLKF SPTEITAPTM PLDCANASNV 60
Vict/96-R 1 ...... 60
Yama/08-S 1 ...... Q...... I...... 60
Vict/06-S 61 QAVNRSATKG VTLLLPEPEW TYPRLSCPGS TFQKALLISP HRFGETKGNS APLIIREPFI 120
Vict/96-R 61 ...... 120
Yama/08-S 61 ...... A...... 120
Vict/06-S 121 ACGPNECKHF ALTHYAAQPG GYYNGTRGDR NKLRHLISVK LGKIPTVENS IFHMAAWSGS 180
Vict/96-R 121 ....K...... E.K ...... 180
Yama/08-S 121 ....T...... E...... 180
Vict/06-S 181 ACHDGKEWTY IGVDGPDNNA LLKVKYGEAY TDTYHSYANK ILRTQESACN CIGGNCYLMI 240
Vict/96-R 181 .....R...... S.. ...I...... N ...... D..... 240
Yama/08-S 181 ...... S.. ...I...... KN ...... D..... 240
Vict/06-S 241 TDGSASGVSE CRFLKIREGR IIKEIFPTGR VKHTEECTCG FASNKTIECA CRDNSYTAKR 300
Vict/96-R 241 ...... I...... E...... 300
Yama/08-S 241 ...P...I...... 300
Vict/06-S 301 PFVKLNVETD TAEIRLMCTD TYLDTPRPND GSITGPCESN GDKGSGGIKG GFVHQRMESK 360
Vict/96-R 301 ...... E ...... D...... A.. 360
Yama/08-S 301 ...... E ...... D ...... A.. 360
163 | P a g e
Vict/06-S 361 IGRWYSRTMS KTERMGMGLY VKYDGDPWAD SDALAFSGVM VSMKEPGWYS FGFEIKDKKC 420
Vict/96-R 361 ...... K...... T...... L...... E...... 420
Yama/08-S 361 ...... K...... T. .E...L...... E...... 420
Vict/06-S 421 DVPCIGIEMV HDGGKETWHS AATAIYCLMG SGQLLWDTVT GVDMAL 466
Vict/96-R 421 ...... K...... 466
Yama/08-S 421 ...... T...... 466
6. Sequence alignments of H3 genes of influenza H3N2 strains
07/-S 1 QKLPGNDNSTATLCLGHHAVPNGTIVKTITNDQIEVTNATELVQSSSTGEICDSPHQILD 60
07/-R 17 ...... 76
07/-S 61 GENCTLIDALLGDPQCDGFQNKKWDLFVERSKAYSNCYPYDVPDYASLRSLVASSGTLEF 120
07/-R 77 ...... 136
07/-S 121 NNESFNWTGVTQNGTSSACIRRSNNSFFSRLNWLTHLKFKYPALNVTMPNNEKFDKLYIW 180
07/-R 137 ...... S...... 196
07/-S 181 GVHHPGTDNDQIFPYAQASGRITVSTKRSQQTVIPNIGSRPRVRNIPSRISIYWTIVKPG 240
07/-R 197 ...... L...... 256
07/-S 241 DILLINSTGNLIAPRGYFKIRSGKSSIMRSDAPIGKCNSECITPNGSIPNDKPFQNVNRI 300
07/-R 257 ...... 316
07/-S 301 TYGACPRYVKQNTLKLATGMRNVPEKQTR 329
07/-R 317 ...... 345
164 | P a g e
165 | P a g e