Identification of Residues Critical for FPR2 Activation by the Cryptic Peptide Mitocryptide-2 Originating from the Mitochondrial

Published March 22, 2019, doi:10.4049/jimmunol.1900060 The Journal of Immunology

Identiﬁcation of Residues Critical for FPR2 Activation by the Cryptic Peptide Mitocryptide-2 Originating from the Mitochondrial DNA–Encoded Cytochrome b

Simon Lind,* Michael Gabl,*,1 Andre´ Holdfeldt,* Jonas Ma˚rtensson,* Martina Sundqvist,* Kodai Nishino,† Claes Dahlgren,* Hidehito Mukai,† and Huamei Forsman*

Similar to bacteria, synthesis of mitochondrial DNA–encoded proteins requires an N-formylated methionine to initiate translation. Thus, the N-formylated methionine peptides originating from mitochondria should be recognized as danger signals. To date, only one such peptide, denoted as mitocryptide-2 (MCT-2), originating from the N-terminal of the mitochondrial cytochrome b, has been isolated from mammalian tissues. Human neutrophils express FPR1 and FPR2 that detect formyl peptides, and the precise structural determinants for receptor recognition remain to be elucidated. MCT-2 is known to activate neutrophils through FPR2 but not FPR1. The aim of this study was to elucidate the structural determinants of importance for receptor preference and human neutrophil activation in MCT-2 by generating a series of MCT-2 variants. We show that there is an absolute requirement for the N-formyl group and the side chain of Met1 at position 1 of MCT-2 but also the C terminus is of importance for MCT-2 activity. We also uncovered individual side chains that positively contribute to MCT-2 activity as well as those suppressed in the response. The MCT-2 peptide and its two polymorphic variants ([Thr7]MCT-2 and [Ser8]MCT-2) all activated neutrophils, but MCT-2 containing Ile7 and Asn8 was the most potent. We also show that some peptide variants displayed a biased FPR2-signaling property related to NADPH oxidase activation and b-arrestin recruitment, respectively. In conclusion, we disclose several critical elements in MCT-2 that are required for neutrophil activation and disclose structural insights into how FPR2 recognition of this mitochondrial DNA–derived peptide may increase our understanding of the role of FPR2 in aseptic inﬂammation. The Journal of Immunology, 2019, 202: 000–000.

he innate immune system is designed to detect not only the structural phospholipid cardiolipin, ATP, DNA with speciﬁc invading microbial pathogens, but recent studies have also methylation characteristics, and peptides/proteins with a formylated T clearly elucidated that many endogenous damage-associated Met (fMet) at N terminus, originate from mitochondria (3). The molecular patterns (DAMPs) released from necrotic cells are mitochondrial N-formyl peptides released from damaged cells upon

by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. recognized by this defense system (1, 2). Pathogen-associated tissue injury are recognized by a group of G protein–coupled molecular patterns and DAMPs are conserved structural entities receptors, the formyl peptide receptors (FPRs), expressed by proof microbial and endogenous origin, respectively, recognized fessional phagocytes (4, 5). The rationale for the FPRs to recognize by discrete innate immune receptors. Many DAMPs, including mitochondrial DNA–derived formyl peptides upon cell damage is that this bacteria-originating organelle starts/initiates the translation of the 13 DNA-encoded proteins with speciﬁc methylation char- *Department of Rheumatology and Inﬂammation Research, University of Gothenburg, acteristics with an N-formylated fMet (N-fMet) (6, 7), and FPRs, 413 46 Gothenburg, Sweden; and †Laboratory of Peptide Science, Graduate School of Bio-Science, Nagahama Institute of Bio-Science and Technology, Nagahama, Shiga thus, have the capacity to recognize pathogen-associated molecular 526-0829, Japan patterns as well as DAMPs. Accordingly, mitochondrial DNA–

http://classic.jimmunol.org 1Current address: Sahlgrenska Cancer Center, University of Gothenburg, Gothenburg, derived formyl peptides with proinflammatory activity have been Sweden. shown to attract neutrophils and to be present in peripheral blood, ORCIDs: 0000-0001-6469-4345 (S.L.); 0000-0002-1988-9247 (J.M.); 0000-0002- both from patients with sepsis and severe aseptic trauma (8, 9). 0859-0792 (M.S.). Among the large number of peptides that can be generated Received for publication January 17, 2019. Accepted for publication February 25, by proteolytic cleavage of the 13 mitochondrial DNA–encoding 2019. proteins, mitocryptide-2 (also known as cryptic mitochondrial This work was supported by the Swedish Research Council, the King Gustaf V 80- Downloaded from Year Foundation, the Swedish government under the Avtal om la¨karutbildning peptide 2; MCT-2) with a sequence identical to the 15 aa present och forskning–agreement, the Clas Groschinsky Memorial Foundation, the in the N terminus of the mitochondrial DNA–encoded cytochrome IngaBritt and Arne Lundberg Foundation, the Wilhelm and Martina Lundgrens Sci- b, has been purified from mammalian tissues and shown to be a entific Foundation, and the Radman˚ and Mrs. Ernst Collianders Foundation. The present work was also supported by a research grant from the Ministry of Education, high-affinity agonist with specificity for FPR2, one of the two Culture, Sports, Science and Technology of Japan (25350971). FPRsexpressedbyneutrophils(10–12).Thatis,inadditionto Address correspondence and reprint requests to Dr. Huamei Forsman, Department of FPR2, human neutrophils also express another FPR member Rheumatology and Inflammation Research, University of Gothenburg, Guldhedsgatan (FPR1), and these two receptors share 69% similarity at the amino 10A, 413 46 Gothenburg, Sweden. E-mail address: [email protected] acid level and are both involved in host defense and the regulation Abbreviations used in this article: CHO, Chinese hamster ovary; CL, chemiluminescence; CysH, cyclosporin H; DAMP, damage-associated molecular pattern; fMet, of inflammation (13, 14). The ligand-binding profile for the neu- formylated Met; FPR, formyl peptide receptor; KRG, Krebs–Ringer phosphate buffer trophil FPRs is complex, as both receptors display high affinity for containing glucose; MCT-2, mitocryptide-2; N-fMet, N-formylated fMet; TFE, 2,2,2- peptides with an fMet at the N terminus but with somewhat dif- trifluoroethanol. ferent ligand preferences with a molecular background not yet Copyright Ó 2019 by The American Association of Immunologists, Inc. 0022-1767/19/$37.50 clearly understood (4). FPR1 is a high-affinity receptor for fairly

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1900060 2 STRUCTURE–ACTIVITY RELATIONSHIP OF MITOCRYPTIDE MCT-2

short bacteria-derived formyl peptides, whereas FPR2 specifically hypotonic lysis of remaining erythrocytes, the neutrophils obtained by recognizes formylated PSMs (a‐type and b‐type phenol soluble centrifugation in a Ficoll-Paque gradient were washed twice in KRG. 3 7 modulins, consisting of 20–26 or ∼44 residues, respectively) secreted The cells were resuspended in KRG (1 10 /ml) and stored on ice until use. by community-associated methicillin-resistant Staphylococcus aureus strains (15, 16). In addition to the formyl peptides, both FPRs also Measurement of superoxide anion production recognize a number of nonformylated molecules, including lipopeptides The production of superoxide anion by the neutrophil NADPH oxidase and peptidomimetics (17, 18). The precise structure present in the was measured by isoluminol-amplified chemiluminescence (CL) in a six- formyl peptides that determines FPR1 versus FPR2 binding is not channel Biolumat LB 9505 (BERTHOLD TECHNOLOGIES, Wildbad, known, but current structure–activity relationship data obtained with Germany) as described earlier (23, 24). In short, 2 3 105/ml neutrophils were mixed (in a total volume of 900 ml) with HRP (4 U) and isoluminol bacteria-derived formyl peptide variants suggests that both peptide 2 (6 3 10 5 M) in KRG and preincubated at 37˚C for 5 min, after which size and amino acid composition are important determinants for re- the stimulus (100 ml) was added. The light emission was recorded con- ceptor preference (17, 19, 20). The importance of FPRs not only in tinuously. When required, the specific receptor inhibitors were included in host defense but also in the regulation of aseptic inflammation imply the CL mixture during the 5 min preincubation at 37˚C before stimulation. that knowledge at the molecular level of FPR-based recognition of b-Arrestin recruitment assay mitochondrial DNA–derived formyl peptides has direct clinical relevance and can be of help when designing novel FPR-targeting The ability of FPR2 agonists to promote b-arrestin recruitment was molecules for treatment of inflammatory-associated diseases. evaluated in the PathHunter eXpress system with CHO-K1 FPR cells and In this study, we set out to investigate the structural requirement accompanying reagents from DiscoverX, as described previously (25). In brief, CHO cells overexpressing FPR2 or FPR1 were diluted in Cell for MCT-2–induced neutrophil response using a series of MCT-2 Plating Reagent before being seeded in tissue culture–treated 96-well variant peptides (including the natural occurring polymorphism plates (10,000 cells per well) and incubated for 20 h (37˚C, 5% CO2). variants) to determine the relation between peptide structure and Cells were then stimulated with agonists for 90 min at 37˚C followed by the neutrophil activation potency and receptor preference. The the addition of detection solution for 60 min, and CL was measured using a Mitras LB 940 plate reader (BERTHOLD TECHNOLOGIES), or activation potencies were evaluated in primary neutrophils, and a CLARIOstar platereader (BMG Labtech). FPR2 preference was confirmed in cultured Chinese hamster ovary (CHO) cells overexpressing individual FPRs. Our data Circular dichroism spectroscopy reveal that the N-fMet residue is a prerequisite for activation/ Circular dichroism spectra of the MCT-2 peptides were obtained at 25˚C recognition and that the peptide size matters for MCT-2 activity. using a JASCO J-820 spectrometer (JASCO) in a quartz cell with a 0.1-cm An Ala scanning approach revealed several key residues that path length. Spectra were collected between 190 and 250 nm with a scan either contribute to or suppress the MCT-2 activation potency. speed of 50 nm/min and a bandwidth of 1 nm. Peptide solutions with a m In addition, MCT-2 containing Ile7 and Asn8 was found to be final concentration of 100 M were prepared in 10 mM phosphate buffer (pH 7.4) with or without 50% 2,2,2-trifluoroethanol (TFE). The baseline the most active MCT-2 peptide among the possible naturally scan, which was acquired by measuring the buffer alone, was subtracted occurring polymorphic variants. Finally, a close comparison of from the experimental readings. Circular dichroism data, which were col- the potencies for NADPH oxidase activation and b-arrestin re- lected every 1 nm, were the average of five scans. The results were expressed cruitment revealed peptides possessing a signaling bias down- as the optical rotation (millidegrees).

by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. stream of FPR2. Data collection and statistical analysis Materials and Methods Data analysis was performed using GraphPad Prism 8 (GraphPad Software, San Diego, CA). EC50 values for MCT-2 derivatives were determined by Ethics statement fitting a sigmoidal dose-response curve with variable slope to data normalized to the maximal response (100%) using the “log (agonist) versus This study, conducted at the Sahlgrenska Academy in Sweden, includes normalized response” in GraphPad Prism. All statistical analysis was blood from buffy coats of healthy individuals obtained from the blood performed on raw data values by using a two-tailed paired Student t test bank at Sahlgrenska University Hospital, Gothenburg, Sweden. According (Fig. 1C) or a repeated measurement of one-way ANOVA followed by a to the Swedish legislation section code 4x 3p SFS 2003:460 (Lag om Dunnett multiple comparison test (Figs. 2C, 3A, 3D). Statistical signi- etikpro¨vning av forskning som avser ma¨nniskor), no ethical approval was ficance is indicated by *p , 0.05, ns: not statistically significant.

http://classic.jimmunol.org needed because the buffy coats were provided anonymously and could not be traced back to a specific donor. Chemicals and reagents Results MCT-2 activates human neutrophils through specific Isoluminol was purchased from Sigma-Aldrich, and HRP was purchased interaction with FPR2 for which the N-terminal formyl group from Roche Diagnostics (Bromma, Sweden). Dextran and Ficoll-Paque were obtained from GE Healthcare Bio-Science (Uppsala, Sweden). The plays an essential role hexapeptide WKYMVM (agonist for FPR2) was from Alta Bioscience, Downloaded from Among the large number of fMet-containing peptides that the- and the fMLF (agonist for FPR1) was purchased from Sigma-Aldrich. Cyclosporin H (CysH; antagonist for FPR1) was kindly provided by oretically can be generated from the 13 proteins encoded for by Novartis Pharma (Basel, Switzerland). The FPR2-specific antagonists the mitochondria DNA, MCT-2 is so far the only one that has been gelsolin-derived inhibitory peptide PBP10 (gelsolin residues 160–169 (21)) isolated from mammals (11), and characterization of this cryptic was from CASLO Laboratory (Lyngby, Denmark). RPMI 1640, FCS, peptide demonstrated MCT-2 to be a potent FPR2-specific penicillin–streptomycin, and G418 were from PAA Laboratories. The neutrophil-activating agonist (10–12). In this study, MCT-2– parent human MCT-2 used (also presented as the MCT-21–15 in this study) was predicted by the homology to the earlier-purified MCT-2 from porcine induced neutrophil activation was assessed by the ability of heart (11). MCT-2 and all MCT-2 derivatives (Table I) were chemically the peptide to initiate a release of superoxide anion generated synthesized as described earlier (10). All peptides were dissolved in by the neutrophil NADPH oxidase (Fig. 1A). In accordance with DMSO, and further dilutions were made in Krebs–Ringer phosphate buffer the earlier-described FPR2 preference in human neutrophils, containing glucose (KRG) (10 mM), Ca2+ (1 mM), and Mg2+ (1.2 mM) (pH 7.3). the MCT-2–induced response could be completely inhibited by the specific FPR2 inhibitor PBP10 but was unaffected by the Isolation of human neutrophils from peripheral blood FPR1-specific inhibitor (Fig. 1A). The prototype FPR2 agonist Blood neutrophils were isolated, as described by Boyum et al. (22), from WKYMVM and FPR1-specific agonist fMLF were included for buffy coats of healthy volunteers. After dextran sedimentation and comparison (Fig. 1B). The Journal of Immunology 3

FIGURE 1. A critical role of the N-terminal formyl group of MCT-2 in activating FPR2-mediated superoxide release from human neutrophils. Neu- trophils were stimulated with MCT-2 or N-terminal–modiﬁed variants, and superoxide release was recorded continuously, expressed as Mega cpm (Mcpm), using the isoluminol/HRP–ampliﬁed CL technique. (A) A representative CL response out of three induced by MCT-2 (500 nM, solid line) is shown. The

FPR2 preference was determined with the FPR1- and FPR2-speciﬁc inhibitors CysH (1 mM, dashed line) and PBP10 (1 mM, dotted line), respectively. (B) The bar graph shows a summary of the inhibitory effect of CysH (1 mM, black) and PBP10 (1 mM, gray) on MCT-2 (500 nM). The prototype FPR2 agonist WKYMVM (100 nM) and the FPR1-speciﬁc agonist fMLF (100 nM) were included for comparison. Data are expressed as remaining activity and calculated from each individual agonist peak response in the presence and absence of the antagonist (n = 3, mean + SD). (C) One representative CL response showing the agonistic effect of the two MCT-2 variants that either lack the formyl group (N-deFor, dashed line, 1 mM) or with the formyl group replaced by an acetyl group (N-acetyl, dotted line, 1 mM) which was determined by stimulating neutrophils with these peptide variants. The FPR2 agonist WKYMVM (100 nM) was used as a positive control (solid line). Inset: The antagonistic effect of MCT-2 was determined by preincubating neutrophils with the peptide variants (1 mM) before stimulation with WKYMVM (100 nM). Data are expressed as the remaining activity calculated from the peak values of WKYMVM response in the presence and absence of peptide variants (n = 3, mean + SD).

It is known that the formyl group plays an essential role for both community-acquired methicillin-resistant S. aureus, has been FPR1 and FPR2 to recognize bacteria-derived formyl peptides shown to critically depend not only on the N terminus but also on (4). To determine the impact of the formyl group in mediating the C-terminal part of the peptide (19). To evaluate the importance host-derived formyl peptides, we designed an MCT-2 variant that of the C terminus of MCT-2 for activation and receptor preference, lacks the formyl group on the N-terminal Met, and this peptide we synthesized a series of stepwise truncations of the parent MCT-2 was not able activate neutrophils when tested in concentrations (also referred as MCT-21–15; Table I). Comparing the full-length up to 2 mM (Fig. 1C). Further, no antagonist effects were ob- peptide MCT-21–15, the MCT-21–14 peptide lacking the last Asn tained with the nonactivating MCT-2 peptides in which the displayed a 3-fold reduced potency in activating the NADPH formylgroupwaseitherremovedorreplacedbyanacetylgroup oxidase–mediated superoxide release (Fig. 2A, 2B), whereas the (N-acetyl MCT-2) (Fig. 1C Inset). Taken together, these data full response induced by the two peptides was similar. Removal by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. show that the presence of the formyl group in the N terminus is also of Ile in position 14, generating the peptide MCT-21–13, an essential requirement for the ability of MCT-2 to interact with resulted in further reduction of potency regarding the triggering neutrophils. of reactive oxygen species release (Fig. 2C). Further stepwise truncations from the C terminus, generating even shorter peptides The C terminus of MCT-2 is of importance for (MCT-21–12, MCT-21–11, and MCT-21–10), resulted in peptides neutrophil activation with much weaker activities compared with MCT-21–15 (Fig. 2C), The size of the proposed receptor-binding pocket in FPRs avail- and peptides shorter than 10 aa residues were completely devoid able for fMet-containing agonists has been estimated to be limited of activity (data not shown).

http://classic.jimmunol.org and consist of four, or at the most, 5 aa (26). Despite this, the Taken together, our data obtained with the N-terminal defor- receptor preference and neutrophil activation potency of the mylated and the progressively C-terminal MCT-2 truncations PSMa2 peptide, a 21-aa-long FPR2-speciﬁc agonist secreted from demonstrate that not only the formyl group in the N terminus but

Table I. MCT-2 truncations

Downloaded from Amino Acid Sequence (Position 1–15)

Peptide 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

MCT-21–15 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Asn MCT-21–14 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile MCT-21–13 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu MCT-21–12 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys MCT-21–11 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met MCT-21–10 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu MCT-21–9 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro MCT-21–8 Formyl-Met Thr Pro Met Arg Lys Ile Asn MCT-21–7 Formyl-Met Thr Pro Met Arg Lys Ile MCT-21–6 Formyl-Met Thr Pro Met Arg Lys MCT-21–5 Formyl-Met Thr Pro Met Arg MCT-21–4 Formyl-Met Thr Pro Met

The table displays amino acid sequences of the formylated MCT-2 and its C-terminal truncated derivatives used in the study. MCT-21–15 with 15 residues long is the parental peptide. 4 STRUCTURE–ACTIVITY RELATIONSHIP OF MITOCRYPTIDE MCT-2

FIGURE 2. Impact of the C terminus of MCT-2 on NADPH oxidase activation and granule secretion. Neutrophils were activated with MCT-2 C terminus

truncations, and dose-dependent release of superoxide was recorded upon stimulation with different peptide concentrations. EC50 values were determined from normalized values to the maximal peak response (100%) from concentration titrations of (A) MCT-21–15 and (B) MCT-21–14 and are presented with a ﬁtted curve (n = 4, mean 6 SD), and concentrations of agonist producing 50% response are indicated by the dashed lines. (C) Neutrophils were stimulated with 1 mM of MCT-2 variants containing 10–14 aa residues that were truncated from the C terminus. Superoxide release was recorded from three in-

dependent experiments, and the peak superoxide response for each peptide variant is presented as the percentage of the full-length MCT-21–15 (1 mM, n =3, mean + SD). *p , 0.05.

also the C-terminal amino acids are of importance for MCT-2– substitution (Fig. 3B). Peptides ([Ala5]MCT-2, [Ala6]MCT-2, induced activity. [Ala8]MCT-2, [Ala11]MCT-2, [Ala12]MCT-2, and [Ala15]MCT-2) with Ala substitution at position 5, 6, 8, 11, 12, and 15 activated Identiﬁcation of multiple MCT-2 residues that contribute to neutrophils at a similar extent to MCT-2 (Fig. 3A). Ala substitu- neutrophil activation tions at position 2, 3, or 9 ([Ala2]MCT-2, [Ala3]MCT-2, and [Ala9] Next, we applied an Ala substitution-scanning approach to gain MCT-2) resulted in an increased superoxide production when structure–activity insights of the importance of individual side 1-mM peptides were compared with MCT-2 (Fig. 3A–C). chains of amino acid residues in MCT-2. We sequentially replaced Interestingly, our data show that the “wild type” MCT-21–15 is each individual amino acid with Ala (Table II), and the potencies of not the most potent neutrophil-activating peptide; the peptide 2 2 these peptides to activate neutrophils were determined. As expected, [Ala ]MCT-21–15 with an Ala substitution of Thr , as well as that the [Ala9]MCT-2 , with an Ala substitution of Pro9, were more

by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. a replacement of the N-terminal fMet residue with formylated Ala 1–15 1 (fMet →formyl-Ala) resulted in abolished capacity to activate potent, with EC50 values of 9 and 43 nM, respectively (Fig. 3C). neutrophils (Fig. 3A). Peptides with individual Ala substitution at To further examine the importance of Thr2 and Pro9 for MCT-2– 4 7 10 residue 4, 7, 10, 13, or 14 ([Ala ]MCT-2, [Ala ]MCT-2, [Ala ] induced neutrophil activation, we designed MCT-21–10 Ala sub- 13 14 2 9 MCT-2, [Ala ]MCT-2, or [Ala ]MCT-2), all had decreased abil- stitutions ([Ala ]MCT-21–10 and [Ala ]MCT-21–10). Similar to the 9 ity to generate superoxide when the activities from a 1-mMcon- inactive MCT-21–10 (Fig. 2C), [Ala ]MCT-21–10 did not activate 2 centration of each peptide were compared (Fig. 3A). The reduced neutrophils to release superoxide, whereas [Ala ]MCT-21–10,in 2 activity of the peptides was also evident when the EC50 values were which Thr was substituted with Ala was active, with an 7 7 compared, as shown for the peptide [Ala ]MCT-2 with an Ile →Ala EC50 value of 470 nM (Fig. 3C, 3D). The maximal activity of http://classic.jimmunol.org

Table II. MCT-2 alanine substitutions

Amino Acid Sequence (Position 1–15)

Peptide 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Downloaded from [Ala1]MCT-2 Formyl-Ala Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala2]MCT-2 Formyl-Met Ala Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala3]MCT-2 Formyl-Met Thr Ala Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala4]MCT-2 Formyl-Met Thr Pro Ala Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala5]MCT-2 Formyl-Met Thr Pro Met Ala Lys Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala6]MCT-2 Formyl-Met Thr Pro Met Arg Ala Ile Asn Pro Leu Met Lys Leu Ile Asn [Ala7]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ala Asn Pro Leu Met Lys Leu Ile Asn [Ala8]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Ala Pro Leu Met Lys Leu Ile Asn [Ala9]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Ala Leu Met Lys Leu Ile Asn [Ala10]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Ala Met Lys Leu Ile Asn [Ala11]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Ala Lys Leu Ile Asn [Ala12]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Ala Leu Ile Asn [Ala13]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Ala Ile Asn [Ala14]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ala Asn [Ala15]MCT-2 Formyl-Met Thr Pro Met Arg Lys Ile Asn Pro Leu Met Lys Leu Ile Ala The table displays amino acid sequences of the MCT-2 alanine substitutions used in the study. Each individual residue in the MCT-2 has been sequentially substituted with Ala (highlighted in bold) from position 1 ([Ala1]MCT-2) to 15 ([Ala15]MCT-2). The Journal of Immunology 5

FIGURE 3. Impact of individual amino acid residue in MCT-2 on neutrophil NADPH oxidase activation. Neutrophils were prewarmed for 5 min at 1 15 37˚C before stimulation with MCT-21–15 alanine substitutions (Ala -Ala ), and superoxide release was measured by isoluminol/HRP–ampliﬁed CL. (A) Summary of superoxide release induced by all 15 Ala substitutions (1 mM). Data are presented as percentage of MCT-2 calculated from the peak 3 CL response, (n = 3, mean + SD). *p , 0.05. (B) Dose responses with given EC50 values of three representative Ala substitution peptides [Ala ]MCT-21–15, 6 7 2 9 [Ala ]MCT-21–15, and [Ala ]MCT-21–15 with increased, unchanged, and reduced activity, respectively, are shown. (C) Contribution of Ala and Ala to 2 9 MCT-2 activity is shown by dose responses of Ala and Ala substitutions. EC50 values were determined from normalized data to the maximal response (100%) of a respective titration of each individual peptide and presented with a ﬁtted curve (n = 3, mean 6 SD). (D) The oxidase activity induced by the 2 9 short derivative MCT-21–10 was compared with [Ala ]MCT-21–10 and [Ala ]MCT-21–10 with alanine substitutions at position 2 and position 9, respectively (n = 3, mean + SD). by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd.

2 [Ala ]MCT-21–10 was comparable to that of the full-length neutrophil response induced by the polymorphic variants was inhibited by FPR2-speciﬁc antagonists (shown with PBP in MCT-21–15 (Fig. 3D inset). 10 Taken together, the data obtained using the Ala screening ap- Fig. 4B) but unaffected by the FPR1 antagonist CysH (Fig. 4B), suggesting that although these two polymorphic variants are less proach reveal an essential role for the MCT-21–15–induced activation of neutrophils of the N-terminal fMet; a dramatic reduction potent than the parental MCT-21–15, they prefer FPR2 over FPR1. in activity was also obtained with the Ala substitution peptides Taken together, these data demonstrate that all natural MCT-2 4 7 10 13 variants are recognized by FPR2 and activate neutrophils, but ([Ala ]MCT-21–15, [Ala ]MCT-21–15, [Ala ]MCT-21–15, [Ala ] 14 http://classic.jimmunol.org they differ in activation capacity, with the parental peptide MCT-21–15, and [Ala ]MCT-21–15), in which the amino acid in position 4, 7, 10, 13, and, 14 was replaced by an Ala, whereas the MCT-21–15 being the most potent. 2 neutrophil activity triggered by the peptides [Ala ]MCT-21–15 and [Ala9]MCT-2 , in which the amino acids in position 2 and 9 All neutrophil-activating MCT-2 peptides promote recruitment 1–15 b were replaced by Ala was increased. of -arrestin with FPR2 as the preferred receptor Many G protein–coupled receptors, including FPR2, have an FPR2 is also the preferred receptor for natural-occurring Downloaded from 7 8 ability to transduce biased signals that can selectively activate [Thr ]MCT-2 and [Ser ]MCT-2 variants, but their activating one cellular response but not another, a feature termed biased capacities are reduced compared with MCT-2 containing 7 8 agonism/signaling or functional selectivity (25, 27). Depending on Ile and Asn the nature of earlier-described FPR2 agonist, b-arrestin is The parental MCT-21–15 used in this study has a peptide sequence recruited to the activated receptor or not (25). The ability of containing Ile7 and Asn8 (formyl-MTPMRKINPLMKLIN), a se- the different MCT-2 variants to trigger b-arrestin recruitment quence that corresponds to the N-terminal of one of the poly- was evaluated using CHO cells expressing b-arrestin tagged with morphic variants of mitochondrial DNA–encoded cytochrome b a b-galactosidase enzyme fragment (EA) and FPR2 tagged with (UniprotKB – P00156). There are two more polymorphic MCT-2 a complementary part of the b-galactosidase enzyme (PK); variants in which the amino acids in position 7 and 8 (VAR008585 b-arrestin recruitment to the activated receptor forces the enzyme and VAR013643) differ from the one used in this study. When the fragments together to form a functional enzyme (25). The activity of three MCT-2 variants was compared, we found that neutrophil-activating MCT-2 peptides recruited b-arrestin 7 8 [Thr ]MCT-21–15 and [Ser ]MCT-21–15 variants were less potent (Fig. 5A–C, Table III), and the relative potencies of the different than MCT-2 containing Ile7 and Asn8 in activating neutrophils peptides were in large agreement with the NADPH oxidase acti- (Fig. 4A). Similar to MCT-2 containing Ile7 and Asn8,the vation proﬁle. The peptides in which the formyl group in the N 6 STRUCTURE–ACTIVITY RELATIONSHIP OF MITOCRYPTIDE MCT-2

FIGURE 4. Neutrophil NADPH oxidase activation by the two polymorphic MCT-2 variants at Thr7 and Ser8.(A) Neutrophils were activated with different concentrations of polymorphic MCT-2 variants at Thr7 (triangle) and Ser8 (closed circle) as indicated, and the release of superoxide was recorded.

EC50 values are determined from normalized data to the maximal response (100%) and presented with a ﬁtted curve (n = 3, mean 6 SD). (B) The FPR preference of [Thr7]MCT-2 (1 mM) and [Ser8]MCT-2 (1 mM) in human neutrophils was determined by preincubating cells with FPR1-speciﬁc inhibitor

CysH (1 mM, black bars) or FPR2 inhibitor PBP10 (1 mM, gray bars). Superoxide release was recorded, and data are expressed as the remaining activity calculated from the peak values of response in cells activated in the absence of inhibitor (n = 3, mean + SD).

terminus was removed or replaced by acetyl group or the first also in b-arrestin translocation (Table III). From the data pre- Met residue was replaced by an Ala, were all unable to recruit sented, it is clear that the preferred receptor for all b-arrestin b-arrestin in concentrations up to 2 mM (Table III). The peptide recruiting MCT-2 peptides is FPR2 (Fig. 3A, Table III). 7 [Ala ]MCT-21–15, being a weak activator of the NADPH oxidase A biased signaling property for the MCT-2 derivatives was cal- (Fig. 3), was also a poor recruiter of b-arrestin (Table III). The culated based on the EC50 values for activation of the NADPH 2 9 potent NADPH oxidase activating [Ala ]MCT-21–15 and [Ala ] oxidase activation and b-arrestin recruitment as described earlier MCT-21–15 peptides recruited b-arrestin with EC50 values lower (10). The biased signal ratio was very close to one for most of the 3 9 than the original MCT-21–15 (Table III). No activation of the MCT-2 peptides, except [Ala ]MCT-21–15 and [Ala ]MCT-21–15 8 NADPH oxidase was induced by the truncated MCT-21–10, together with the [Ala ]MCT-21–15 (Table III). This suggests that 2 3 9 whereas the [Ala ]MCT-21–10 activated the oxidase (Fig. 3D). these three peptides ([Ala ]MCT-21–15,[Ala]MCT-21–15,and 2 8 Similar to these data, [Ala ]MCT-21–10 but not MCT-21–10 [Ala ]MCT-21–15) are biased agonists either toward b-arrestin 3 9 recruited b-arrestin (Table III). Also, the two natural polymorphic ([Ala ]MCT-21–15) or toward oxidase activation ([Ala ]MCT-21–15 7 8 8 MCT-2 variants ([Thr ]MCT-21–15 and [Ser ]MCT-21–15) dose- and [Ala ]MCT-21–15), respectively. by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. dependently recruited b-arrestin and, compared with the potency Taken together, we show that the b-arrestin recruitment data for of the MCT-21–15, the potencies were somewhat lower (Fig. 5B, the MCT-2 variants are in large agreement with their activating Table III). The two variants with Ala substitutions in position 2 profile in the NADPH oxidase assay, the oxidase-activating variants 2 9 and9([Ala]MCT-21–15 and [Ala ]MCT-21–15), displaying also triggered b-arrestin translocation. However, we also reveal a significantly increased potencies in activation of the NADPH biased FPR2 signaling property for some cryptic mitochondrial oxidase were more potent than the parental MCT-21–15 peptide peptide (MCT) variants in activating the two systems. Further, the http://classic.jimmunol.org Downloaded from

FIGURE 5. Effect of MCT-2 variants in promoting FPR2 to recruit b-arrestin. PathHunter CHO cells that coexpress ProLink-tagged FPR with b-arrestin were challenged with MCT-2 variants as indicated, and b-arrestin translocation to the plasma membrane–bound FPR was measured as b-galactosidase

activity through enzyme fragment complementation. (A) Effect of MCT-21–15 and its C terminus truncations down to 10 aa (1 mM) on arrestin recruitment in CHO cells overexpressing FPR1 (black bars) or FPR2 (gray bars). Data are presented as percentage of response from 100 nM FPR1 agonist fMLF in the FPR1-overexpressing cell line and 100 nM WKYMVM for the FPR2-overexpressing cell line, respectively. Data are presented from separate experiments from different cell passages (n = 3, mean + SD). (B) Dose-dependent effect of MCT-2 polymorphic variants [Thr7]MCT-2 (black bars) and [Ser8]MCT-2

(gray bars) in activating FPR2 cells to recruit b-arrestin. Data are presented as percentage of MCT-21–15 (1 mM) (n = 3, mean + SD). (C) Effect of MCT-2 alanine substitutions Ala2 (black bars) and Ala9 (gray bars) on FPR2-mediated b-arrestin recruitment. Data are expressed as percentage of MCT-2 (1 mM) and presented as mean + SD from three independent experiments on cells from different passages. The Journal of Immunology 7

Table III. Structure–activity relationship of MCT-2 and its derivatives

Neutrophil NADPH Oxidase FPR1/FPR2 b-Arrestin Recruitment a b c Mitocryptide MCT-2 EC50 Value (nM) Preference EC50 (nM)

MCT-21–15 160 (144–176) FPR2 130 MCT-2 N-terminal modiﬁcations [N-Ac]MCT-2 — — — [N-deFor]MCT-2 — — — MCT-2 C-terminal truncations MCT-21–14 420 (387–455) FPR2 180 d d MCT-21–13 .1000 FPR2 500 d d MCT-21–12 .1000 FPR2 .1000 d d MCT-21–11 .1000 FPR2 .1000 MCT-21–10 ——— MCT-21–9 ——— MCT-21–8 ——— MCT-21–7 ——— MCT-21–6 ——— MCT-21–5 ——— MCT-21–4 ——— MCT-21-15 alanine substitutions [Ala1]MCT-2 — — — [Ala2]MCT-2 9 (8–10) FPR2 6 [Ala3]MCT-2 210 (190–230) FPR2 560 [Ala4]MCT-2 .1000d — .1000d [Ala5]MCT-2 .1000d — .1000 [Ala6]MCT-2 488 (451–529) FPR2 [Ala7]MCT-2 1500 (1326–1679) FPR2 .1000d [Ala8]MCT-2 410 (381–442) FPR2 500d [Ala9]MCT-2 43 (39–42) FPR2 10 [Ala10]MCT-2 .1000d FPR2 .1000d [Ala11]MCT-2 521 (483–563) FPR2 [Ala12]MCT-2 200 (177–226) FPR2 [Ala13]MCT-2 .1000d — .1000d [Ala14]MCT-2 .1000d — .1000d [Ala15]MCT-2 408 (358–464) FPR2 MCT-21-10 variants [Ala2]MCT-2 470 (413–532) FPR2 750 [Ala9]MCT-2 — — — MCT-21-15 polymorphic variants [Ser8]MCT-2 550 (491–604) FPR2 250 by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. [Thr7]MCT-2 830 (746–895) FPR2 780

a Neutrophil NADPH oxidase is determined according to the method described in the Materials and Methods section, and the EC50 values are calculated based on data from three independent experiments. Ninety-five percent confidence intervals (asymmetrical) are shown in parentheses. bReceptor preference between the two neutrophil-expressing FPRs (FPR1 and FPR2) is determined by using well-characterized FPR-specific antagonist in combination with FPR1/FPR2–overexpressing CHO cells to study b-arrestin translocation. c b-Arrestin recruitment is determined according to the method described in the Materials and Methods section, and the EC50 values are calculated based on data from three independent experiments. Ninety-five percent confidence intervals (asymmetrical) are shown in parentheses. dData are estimated from two independent experiments. —, no activity could be detected.

http://classic.jimmunol.org arrestin assay using individual FPR overexpressing cells conﬁrmed Discussion FPR2 preference for all peptides. Mitochondria, organelles with an ancestral origin from bacteria, have, in addition to their role in providing the energy requirements Alanine 2 or 9 substitution in MCT21–15 results in an increased for cells and tissues, abilities to regulate the function of our innate a‐ helical structure immune system in combating invading microbes and in the Several of the FPR2-interacting peptides described adopt a‐helical, resolution of sterile insults. Accordingly, mitochondria contains/

Downloaded from amphipathic property (20, 28), suggesting that the a-helical generates endogenous DAMPs that can be released from damaged structure may play an important role for high-afﬁnity interaction cells/tissues and trigger immune/inﬂammatory responses (4, 29). with FPR2. The secondary structures of the parental MCT-21–15 MCT-2, a 15-aa-long peptide originating from the mitochondrial 2 and the two Ala-substituted derivatives, [Ala ]MCT-21–15 and DNA–encoded cytochrome b has been isolated from mammalian 9 [Ala ]MCT-21–15 that had dramatically increased activity, were tissue extract (10, 11, 30), and the structure–activity relationship examined using circular dichroism, an excellent tool for rapid of MCT-2 have been investigated in the current study. We show investigation of the secondary structure. A comparison of the that the formyl group in the N terminus and the amino acids in the circular dichroism revealed that the full‐length MCT-21–15 con- C terminus are of utmost importance for the ability of MCT-2 to tained a-helical structure in membrane-mimicking TFE solution, activate neutrophils. Peptides with individual Ala substitutions, but it has random-coiled structure in aqueous phosphate buffer used to uncover residues important for function, were either silent (Fig. 6). The contents of a-helical structure in TFE buffer were or critical for function and promoted or reduced their agonistic increased by the substitution of Pro9 for Ala in MCT-2 but not that activity. The activation potency also differed between polymor- of Thr2 for Ala. These data indicate that the a-helical structure of phic MCT-2 variants. All activating MTC-2 variant peptides were MCT-2 at its C terminus but not the N-terminal part contribute to recognized by FPR2 but not FPR1, and some MCT-2 variants FPR2 binding and neutrophil activation. transduced a biased signaling through this receptor toward their 8 STRUCTURE–ACTIVITY RELATIONSHIP OF MITOCRYPTIDE MCT-2

oxidase or a recruitment of b-arrestin (10). Although the earlier- described formylated peptides with sequences originating from mitochondrial DNA–encoded proteins activate neutrophils, no such peptides have been isolated/identified in mammalian tissues, except for the MCT-2 that has been isolated from porcine heart (11). With the attempt to determine the structure–activity relationship of mitochondrial DAMPs, the human variant of this MCT-2 peptide has now been studied in more detail. An essential role of the formyl group in MCT-2 activity is in line with the notion that N-fMet, the hallmark of bacteria- and mitochondria- encoded protein synthesis, is recognized by the innate immune system as foreign and by that as a DAMP. Nevertheless, the formyl group is not an absolute requirement for FPR interaction, as many nonformyl peptides can interact with both FPR1 and FPR2 (4). In contrast to the short formyl peptides (3–5 residues long, including fMLF and fMIFL) derived from bacteria that are high-affinity agonists for FPR1 (19), the 15-residues-long MCT-2 is specifically recognized by FPR2. Truncation of even one amino acid from the C terminus of MCT-2 resulted in a reduced activity, suggesting that the peptide size is critical for FPR2 interaction. This is further strengthened by the fact that the MCT-2 derivatives shorter than 10 aa were nonactivating. The C terminus has also been shown to be important for neutrophil activation by the N-fMet containing PSMa2, a 21-aa-long peptide preferentially recognized by FPR2 (19, 20). It should be noticed that for the PSMa2 peptide the C-terminal part is important not only for the activation potency but also for the receptor preference (19). There is, however, no absolute rule related to the link between peptide length and activity of FPR2-activating peptides, illustrated by the fact that a 10-aa-long FPR2-specific pepducin, generated from the third intracellular loop of FPR2, is more potent than the pepducin with a 16-residues-long peptide (31). The similarity between PSMa2 and MCT-2 in FPR2 activation regarding the impact of length was not linked to a similarity when it comes to a changed

by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. receptor preference; the changed receptor preference of shorter PSMa2 peptides (19) was not seen for shorter MCT-2 peptides in human neutrophils at the concentrations examined. When it comes to FPR1, a structural model of the peptide agonist–receptor complex suggests that the transmembrane part of the binding pocket of the receptor binds optimally to a 4-aa peptide. Both virtual and physical data show that C terminus elongation of the FIGURE 6. Circular dichroism spectra of MCT-2 and its Ala-substituted peptide has no effect on FPR1 binding/activity (26). Even if the 2 9 derivatives [Ala ]MCT-2 and [Ala ]MCT-2. The a-helical secondary fMet and the amino acids close to the N terminus are also of prime http://classic.jimmunol.org structure of peptides was indicated by their circular dichroism signals. importance for peptide agonists that activate FPR2, the FPR1 A 2 B The circular dichroism spectra of MCT-2 ( ), [Ala ]MCT-2 ( ), or model is obviously not valid for FPR2, illustrated by the fact that [Ala9]MCT-2 (C) in 10 mM phosphate buffer (pH 7.4) (solid lines) or in the C terminus of MCT-2 is required for activity. Studies of ho- 50% TFE buffer (dashed lines) were collected. Circular dichroism data with peptide solutions (ﬁnal concentration of 100 mM) were the average of mology modeling of FPR2 and peptide docking have shed some ﬁve scans collected every 1 nm. The results are expressed as the optical insights into ligand recognition, and these data suggest that peptide rotation (mdeg). ligands and nonpeptide ligands are differently placed upon inter-

Downloaded from action with FPR2, vertically in extended conformation, whereas the nonpeptides are linked more horizontally (32, 33). In addition, the abilities to activate the superoxide-generating NADPH oxidase or FPR2 residues Asp281 and Arg26 are crucial for binding of peptide recruit b-arrestin. ligands but not for nonpeptide ligands (32, 33). Future crystal Mitochondrial DNA encodes 13 proteins that are translated with structure of FPR2 will undoubtedly increase our understanding of an fMet in their N termini, and a large number of different cryptic this versatile receptor in recognizing an array of ligands belonging peptides can be formed from these peptides and be recognized by to different classes of molecules. our innate immune system (10, 30). In contrast to our knowledge At the receptor level, earlier gain-of-function studies with chi- about bacterial peptides with the same mark of identification, very meric FPR1/FPR2 receptors indicate that three noncontiguous little is known about the recognition profiles of MCTs. We have clusters of amino acids in the first extracellular loop and its adjacent earlier identified six MCTs that potently activated neutrophils with transmembrane domains are important for high-affinity binding of distinct FPR recognition profiles as dual FPR1/2 agonists or rec- fMLF (34, 35). For FPR2, the negatively charged Asp281 has been ognized preferentially by either FPR1 or FPR2. Some of the suggested to be crucial for interaction with formyl peptides that MCTs were also found to be biased-signaling agonists that signal contain a positively charged C terminus (32). In addition, the in favor for an activation of the superoxide-generating NADPH binding pocket of FPR1 is proposed to have a limited depth and The Journal of Immunology 9

by that is capable of accommodating up to five residues (36). The events may, if not prohibited at an early stage, eventually cause binding pocket for FPR2 is not known, but our circular dichroism inflammation-associated tissuedamageandorganfailure. analysis reveals a critical role of the a-helical structure of MCT-2 In summary, our data confirm that MCT-2, an endogenous formyl in FPR2 recognition. In addition, our Ala substitution data, showing methionyl-containing peptide derived from the mitochondria- that a replacement of hydrophobic amino acid residue at posi- encoded cytochrome b, is a high-affinity FPR2 agonist that po- tions 4, 7, 10, 13, or 14 all had reduced activity, suggests that tently activates human neutrophils. Our study provides structural the amphipathic a-helical structure of the C-terminal part of and signaling insights into FPR2 recognition of MCT-2 and its MCT-2–containing hydrophobic amino acid residues determines naturally occurring variants. This may help in the design of FPR2- high-affinity interaction between MCT-2 and FPR2 in human based therapeutics for treating sterile inflammation associated neutrophils. with tissue damage when excessive neutrophil tissue infiltration or The signaling properties of the MCT-2/FPR2 ligand/receptor systemic inflammation is triggered by mitochondrial DNA–derived complex and the potencies of the original MCT-2 peptide and DAMPs, such as the N-formylated peptides. variants thereof were examined to trigger activation of the NADPH b oxidase and -arrestin translocation, respectively, were deter- Acknowledgments mined. The precise physiological role of b-arrestin translocation We thank the members of the Phagocyte Research Group, University of downstream of FPRs is not known, but our recent data using Gothenburg, Sweden for the valuable suggestions provided. FPR2-biased agonists has suggested a role for signaling downstream of b-arrestin in neutrophil chemotactic migration (25). For MCT-2, we have very recently demonstrated that MCT-2 is che- Disclosures motactic and displays a balanced signaling regarding NADPH The authors have no financial conflicts of interest. oxidase activation and b-arrestin recruitment (10). In this study, we identify both balanced and biased agonists among the MCT-2 References b variants. One regulatory role of -arrestin is commonly consid- 1. Kono, H., and K. L. Rock. 2008. How dying cells alert the immune system to ered to be to terminate G protein–dependent signaling of the danger. Nat. Rev. Immunol. 8: 279–289. agonist-occupied receptor, but recent research has added a novel 2. Kolaczkowska, E., and P. Kubes. 2013. Neutrophil recruitment and function in b health and inflammation. Nat. Rev. Immunol. 13: 159–175. role of -arrestin in initiating a G protein–independent signaling 3. Banoth, B., and S. L. Cassel. 2018. Mitochondria in innate immune signaling. pathway (37). The fact that several FPR2 agonists, including Transl. Res. 202: 52–68. MCT-2, trigger b-arrestin recruitment, whereas others lack this 4. Dahlgren, C., M. Gabl, A. Holdfeldt, M. Winther, and H. Forsman. 2016. Basic characteristics of the neutrophil receptors that recognize formylated peptides, a ability, suggests that future work should examine whether the danger-associated molecular pattern generated by bacteria and mitochondria. different agonists trigger unique signals downstream of FPR2. Biochem. Pharmacol. 114: 22–39. 5. Ye, R. D., F. Boulay, J. M. Wang, C. Dahlgren, C. Gerard, M. Parmentier, Such data might shed light on how FPR2 distinguishes host- C. N. Serhan, and P. M. Murphy. 2009. International union of basic and clinical derived and bacteria-derived danger molecules and how they are pharmacology. LXXIII. Nomenclature for the formyl peptide receptor (FPR) able to mediate both proinflammatory and proresolving responses. family. Pharmacol. Rev. 61: 119–161. 6. Adams, J. M. 1968. On the release of the formyl group from nascent protein. The human genome contains several polymorphic MCT-2 var- J. Mol. Biol. 33: 571–589.

by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. iants, and we show that their abilities to activate neutrophils differ 7. Anderson, S., A. T. Bankier, B. G. Barrell, M. H. de Bruijn, A. R. Coulson, somewhat, as illustrated by the fact that an exchange of an Ile in J.Drouin,I.C.Eperon,D.P.Nierlich,B.A.Roe,F.Sanger,etal.1981. Sequence and organization of the human mitochondrial genome. Nature 290: position 7 for Thr gives rise to a peptide with a lower activation 457–465. potency. Whether such differences in MCT-2 or in any of the other 8. McDonald, B., K. Pittman, G. B. Menezes, S. A. Hirota, I. Slaba, C. C. Waterhouse, P. L. Beck, D. A. Muruve, and P. Kubes. 2010. Intravascular danger signals guide 13 proteins encoded for by the mitochondrial DNA have clinical neutrophils to sites of sterile inflammation. Science 330: 362–366. relevance in the development or resolution of sterile inflammatory 9. Zhang, Q., M. Raoof, Y. Chen, Y. Sumi, T. Sursal, W. Junger, K. Brohi, reactions awaits further investigation. It is worth noting that per- K. Itagaki, and C. J. Hauser. 2010. Circulating mitochondrial DAMPs cause inflammatory responses to injury. Nature 464: 104–107. sons carrying the polymorphic MCT-2 variants that have lower 10. Gabl, M., M. Sundqvist, A. Holdfeldt, S. Lind, J. Martensson, K. Christenson, neutrophil recruitment and activation potencies may have a pro- T. Marutani, C. Dahlgren, H. Mukai, and H. Forsman. 2018. Mitocryptides from human mitochondrial DNA-encoded proteins activate neutrophil formyl peptide http://classic.jimmunol.org longed reaction or a reduced wound healing during an aseptic receptors: receptor preference and signaling properties. J. Immunol. 200: 3269–3282. inflammation. We know that inflammation plays a critical role in 11. Mukai, H., T. Seki, H. Nakano, Y. Hokari, T. Takao, M. Kawanami, the pathogenesis of a number of diseases, including cardiovascu- H. Tsukagoshi, H. Kimura, Y. Kiso, Y. Shimonishi, et al. 2009. Mitocryptide-2: purification, identification, and characterization of a novel cryptide that activates lar disease (38), and the recent Canakinumab Anti-Inflammatory neutrophils. J. Immunol. 182: 5072–5080. Thrombosis Outcome Study (CANTOS) trial has provided proof- 12. Seki, T., A. Fukamizu, Y. Kiso, and H. Mukai. 2011. Mitocryptide-2, a of-concept that targeting the inflammatory pathway has a benefi- neutrophil-activating cryptide, is a specific endogenous agonist for formyl- peptide receptor-like 1. Biochem. Biophys. Res. Commun. 404: 482–487.

Downloaded from cial effect on the cardiovascular event rates (39). It is reasonable to 13. Li, Y., and D. Ye. 2013. Molecular biology for formyl peptide receptors in hu- assume that upon tissue/cell damage caused, for example, by a man diseases. J. Mol. Med. (Berl.) 91: 781–789. myocardial infarction, is associated with a release of DAMPs, 14. Weiß, E., and D. Kretschmer. 2018. Formyl-peptide receptors in infection, in- ﬂammation, and cancer. Trends Immunol. 39: 815–829. including the MCT-2 peptide and other formyl peptides originat- 15. Kretschmer, D., A. K. Gleske, M. Rautenberg, R. Wang, M. Koberle, E. Bohn, ing from mitochondria, and that these mediators recruit neutro- T. Schoneberg, M. J. Rabiet, F. Boulay, S. J. Klebanoff, et al. 2010. Human formyl peptide receptor 2 senses highly pathogenic Staphylococcus aureus. Cell phils that inﬁltrate the damaged/necrotic tissue. At such sites, Host Microbe 7: 463–473. these endogenous DAMPs may trigger neutrophils to release toxic 16. Forsman, H., K. Christenson, J. Bylund, and C. Dahlgren. 2012. Receptor- reactive oxygen species and hydrolytic enzymes that could cause dependent and -independent immunomodulatory effects of phenol-soluble modulin peptides from Staphylococcus aureus on human neutrophils are abro- further tissue damage and, thus, release of more DAMPs. Ac- gated through peptide inactivation by reactive oxygen species. Infect. Immun. 80: cordingly, it has been shown that in heart failure, increased pro- 1987–1995. duction of reactive oxygen radicals may directly contribute to 17. He, H. Q., and R. D. Ye. 2017. The formyl peptide receptors: diversity of ligands and mechanism for recognition. Molecules 22: E455. mitochondrial oxidative stress and dysfunction (40). In addition, it 18. Skovbakke, S. L., A. Holdfeldt, H. Forsman, J. Bylund, and H. Franzyk. 2018. has been shown that tissue injury promotes releases of formyl The role of formyl peptide receptors for immunomodulatory activities of anti- microbial peptides and peptidomimetics. Curr. Pharm. Des. 24: 1100–1120. peptides into the circulation with functionally important im- 19. Forsman, H., M. Winther, M. Gabl, S. L. Skovbakke, F. Boulay, M. J. Rabiet, and mune consequences [e.g., activation of neutrophils (9)]. Such C. Dahlgren. 2015. Structural changes of the ligand and of the receptor alters the 10 STRUCTURE–ACTIVITY RELATIONSHIP OF MITOCRYPTIDE MCT-2

receptor preference for neutrophil activating peptides starting with a for- N-formylated peptides including mitocryptide-2 components of mitochondrial mylmethionyl group. Biochim. Biophys. Acta 1853: 192–200. damage-associated molecular patterns? Biopolymers 106: 580–587. 20. Kretschmer, D., M. Rautenberg, D. Linke, and A. Peschel. 2015. Peptide length 31. Forsman, H., J. Bylund, T. I. Oprea, A. Karlsson, F. Boulay, M. J. Rabiet, and and folding state govern the capacity of staphylococcal beta-type phenol-soluble C. Dahlgren. 2013. The leukocyte chemotactic receptor FPR2, but not the modulins to activate human formyl-peptide receptors 1 or 2. J. Leukoc. Biol. 97: closely related FPR1, is sensitive to cell-penetrating pepducins with amino 689–697. acid sequences descending from the third intracellular receptor loop. Biochim. 21. Forsman, H., E. Andreasson, J. Karlsson, F. Boulay, M. J. Rabiet, and Biophys. Acta 1833: 1914–1923. C. Dahlgren. 2012. Structural characterization and inhibitory profile of formyl 32. He, H. Q., E. L. Troksa, G. Caltabiano, L. Pardo, and R. D. Ye. 2014. Structural peptide receptor 2 selective peptides descending from a PIP2-binding domain of determinants for the interaction of formyl peptide receptor 2 with peptide gelsolin. J. Immunol. 189: 629–637. ligands. J. Biol. Chem. 289: 2295–2306. 22. Boyum, A., D. Lovhaug, L. Tresland, and E. M. Nordlie. 1991. Separation of 33. Stepniewski, T. M., and S. Filipek. 2015. Non-peptide ligand binding to the leucocytes: improved cell purity by fine adjustments of gradient medium density formyl peptide receptor FPR2–A comparison to peptide ligand binding modes. and osmolality. Scand. J. Immunol. 34: 697–712. Bioorg. Med. Chem. 23: 4072–4081. 23. Bylund, J., H. Bjornsdottir, M. Sundqvist, A. Karlsson, and C. Dahlgren. 2014. 34. Quehenberger, O., Z. K. Pan, E. R. Prossnitz, S. L. Cavanagh, C. G. Cochrane, Measurement of respiratory burst products, released or retained, during activa- and R. D. Ye. 1997. Identification of an N-formyl peptide receptor ligand binding tion of professional phagocytes. Methods Mol. Biol. 1124: 321–338. domain by a gain-of-function approach. Biochem. Biophys. Res. Commun. 238: 24. Dahlgren, C., A. Karlsson, and J. Bylund. 2007. Measurement of respiratory 377–381. burst products generated by professional phagocytes. Methods Mol. Biol. 412: 35. Quehenberger, O., E. R. Prossnitz, S. L. Cavanagh, C. G. Cochrane, and 349–363. R. D. Ye. 1993. Multiple domains of the N-formyl peptide receptor are required 25. Gabl, M., A. Holdfeldt, M. Sundqvist, J. Lomei, C. Dahlgren, and H. Forsman. for high-affinity ligand binding. Construction and analysis of chimeric N-formyl 2017. FPR2 signaling without beta-arrestin recruitment alters the functional peptide receptors. J. Biol. Chem. 268: 18167–18175. repertoire of neutrophils. Biochem. Pharmacol. 145: 114–122. 36. Sklar, L. A., S. P. Fay, B. E. Seligmann, R. J. Freer, N. Muthukumaraswamy, and 26. Movitz, C., L. Brive, K. Hellstrand, M. J. Rabiet, and C. Dahlgren. 2010. The H. Mueller. 1990. Fluorescence analysis of the size of a binding pocket of a annexin I sequence gln(9)-ala(10)-trp(11)-phe(12) is a core structure for inter- peptide receptor at natural abundance. Biochemistry 29: 313–316. action with the formyl peptide receptor 1. J. Biol. Chem. 285: 14338–14345. 37. Luttrell, L. M., and D. Gesty-Palmer. 2010. Beyond desensitization: physio- 27. Pupo, A. S., D. A. Duarte, V. Lima, L. B. Teixeira, E. S. L. T. Parreiras, and logical relevance of arrestin-dependent signaling. Pharmacol. Rev. 62: 305–330. C. M. Costa-Neto. 2016. Recent updates on GPCR biased agonism. Pharmacol. 38. Ridker, P. M. 2014. Targeting inflammatory pathways for the treatment of car- Res. 112: 49–57. diovascular disease. Eur. Heart J. 35: 540–543. 28. Fu, H., A. Bjorstad, C. Dahlgren, and J. Bylund. 2004. A bactericidal cecropin-A 39. Ridker, P. M., B. M. Everett, T. Thuren, J. G. MacFadyen, W. H. Chang, peptide with a stabilized alpha-helical structure possess an increased killing C. Ballantyne, F. Fonseca, J. Nicolau, W. Koenig, S. D. Anker, et al. 2017. capacity but no proinflammatory activity. Inflammation 28: 337–343. Antiinflammatory therapy with canakinumab for atherosclerotic disease. 29. Jin, H. S., H. W. Suh, S. J. Kim, and E. K. Jo. 2017. Mitochondrial control of N. Engl. J. Med. 377: 1119–1131. innate immunity and inflammation. Immune Netw. 17: 77–88. 40. Kiyuna, L. A., E. A. R. Prestes, C. H. Chen, D. Mochly-Rosen, and J. C. 30. Marutani, T., T. Hattori, K. Tsutsumi, Y. Koike, A. Harada, K. Noguchi, Y. Kiso, B. Ferreira. 2018. Targeting mitochondrial dysfunction and oxidative stress in and H. Mukai. 2016. Mitochondrial protein-derived cryptides: are endogenous heart failure: challenges and opportunities. Free Radic. Biol. Med. 129: 155–168. by guest on October 1, 2021. Copyright 2019 Pageant Media Ltd. http://classic.jimmunol.org Downloaded from