Expression of the Human Poliovirus Receptor/CD155 Gene During Development of the Central Nervous System: Implications for the Pathogenesis of Poliomyelitis

Virology 273, 248–257 (2000) doi:10.1006/viro.2000.0418, available online at http://www.idealibrary.com on

Expression of the Human Poliovirus Receptor/CD155 Gene during Development of the Central Nervous System: Implications for the Pathogenesis of Poliomyelitis

Matthias Gromeier,1 David Solecki, Dhavalkumar D. Patel,* and Eckard Wimmer

Department of Molecular Genetics and Microbiology, School of Medicine, State University of New York at Stony Brook, Stony Brook, New York 11790; and *Department of Medicine, Division of Rheumatology, Allergy and Clinical Immunology, Duke University Medical Center, Durham, North Carolina 27710 Received March 2, 2000; returned to author for revision April 24, 2000; accepted May 9, 2000

The gene for the human poliovirus receptor (hPVR/CD155) is the founding member of a new family of genes encoding proteins belonging to the immunoglobulin superfamily. To determine whether CD155 is expressed during mammalian development, we have made use of the previously characterized promoter of the CD155 gene and generated mice transgenic for a CD155 promoter-driven ␤-galactosidase reporter gene. Expression of the reporter gene in transgenic embryos was observed during midgestation in anterior midline structures of the developing central nervous system and in the neuroretina. During that period, reporter gene expression appeared within the notochord and floor plate along the entire spinal cord reaching into the caudal diencephalon. In addition, transgene expression was observed in axonal projections emanating from retinal ganglion cells forming the optic nerve to reach the future region of the optic chiasm. Analysis of expression of CD155 during human embryonic development confirmed the distribution of reporter gene expression specified by CD155 promoter activity. The anatomical distribution of CD155 promoter activity during embryogenesis matches that of transacting factors previously identified to regulate transcription of the CD155 gene. Expression of CD155 within embryonic structures giving rise to spinal cord anterior horn motor neurons may explain the restrictive host cell tropism of poliovirus for this cellular compartment of the CNS. © 2000 Academic Press

INTRODUCTION parts mediate poliovirus infectivity (Koike et al., 1992). The human poliovirus receptor-related genes 1 (hPRR1; CD155 (also known as hPVR) is the term for four isotypes of glycoproteins belonging to the immunoglob- Lopez et al., 1995) and 2 (hPRR2; Eberle et al., 1995) as ulin superfamily (IGSF) that have been identified as the well as the rodent molecules mPRR2 [previously referred receptors for poliovirus (Mendelsohn et al., 1989). Its to as mouse poliovirus receptor homolog, MPH (Morri- gene is the founding member of a new family of primate son and Racaniello, 1992)] and mTage4 (Chadeneau et and rodent genes that encode polypeptides with a com- al., 1994) lack poliovirus binding activity. Interestingly, mon structural arrangement of three extracellular (V-C2- hPRR1 and hPRR2 have been found to serve as recep- C2) domains (Fig. 1A; Mendelsohn et al., 1989; Koike et tors for ␣-herpesviruses (Geraghty et al., 1998) and pseu- al., 1990; Wimmer et al., 1993). CD155 is expressed in dorabies virus (Warner et al., 1998), respectively. four isoforms: CD155␣ and -␦ are membrane-bound vari- It is believed that CD155 may play a critical role in the ants that differ only in the sequence of their cell-internal determination of specific neuropathogenicity of poliovi- C-terminal domain, while CD155␤ and -␥ are secreted rus that limits susceptibility to spinal and brainstem mo- isoforms lacking the transmembrane domain (Fig. 1; tor neurons (Wimmer et al., 1993; Gromeier et al., 1995). Koike et al., 1990; Wimmer et al., 1993). CD155␣ and -␦ It has been difficult, however, to corroborate this as- are type Ia single-pass transmembrane glycoproteins sumption. Attempts to match the anatomical distribution Ͼ with apparent M r of 80,000 (the core polypeptides are of CD155 with susceptibility to poliovirus failed because 42.5 and 40 kDa, respectively; Bernhardt et al., 1994a). of the difficulty in detecting CD155 expression within Binding of poliovirus occurs at the V-domain of the adult primate neural tissues. Available evidence sug- polypeptide (Bernhardt et al., 1994b; Belnap et al., 2000; gests expression levels of these proteins to be exceed- He et al., 2000). ingly low, both in human tissues and in human tissue Only the CD155 isoforms and their simian counter- culture cell lines (Bernhardt et al., 1994a; Solecki et al., 1997).

1 Mammalian, avian, and insect cell adhesion mole- To whom correspondence and reprint requests should be ad- cules belonging to the IGSF are expressed in key struc- dressed at present address: Department of Microbiology, Box 3020, Duke University Medical Center, Durham, NC 27710. Fax: (919) 684 tures exerting developmental cues within the embryonic 8735. E-mail: [email protected]. CNS (Colamarino and Tessier-Lavigne, 1995; Tessier-

Lavigne and Goodman, 1996; Yoshihara et al., 1991). Rossant, was tested to ascertain that no cryptic expres- Similarity of CD155 with insect IGSF molecules that play sion of the reporter gene occurs after transfection into a role in nervous system development, e.g., Drosophila tissue culture cells (data not shown). As was reported amalgam (Hellen et al., 1989; Seeger et al., 1988), grass- previously, the gene for CD155 is expressed in most hopper lachesin (Karlstrom et al., 1993), or Drosophila human tissue culture cell lines tested with the notable irreC (Ramos et al., 1993), prompted us to investigate the exception of some lymphatic cell lines, including Raji expression of CD155 during embryonic development. cells (an Epstein–Barr virus-transformed lymphoblas- We have studied first the basic structure of the regu- toma cell line). The CD155 promoter is silent in Raji cells latory element of the CD155 gene and identified trans- (Solecki et al., 1997, 1999). The CD155/LacZ vector was acting factors likely to be involved in the regulation of the therefore analyzed for activity in Ntera-2, a human ter- CD155 gene (Solecki et al., 1997, 1999, 2000). This has atocarcinoma cell line expressing CD155, as well as in allowed us to construct a vector in which expression of Raji cells. After transfection of vector DNA, ␤-galactosi- a reporter gene (␤-galactosidase; LacZ) is controlled by dase activity was observed in Ntera-2 cells but not in Raji the CD155 upstream region. Since nonprimate homologs cells (Fig. 2). This observation is an indication that the to CD155 have not been identified, experimental animal regulation of reporter gene expression in the CD155/ models to analyze expression during embryonic devel- LacZ vector may correlate with that of the endogenous opment are not available. Therefore, we used a CD155/ cellular CD155 promoter. lacZ reporter gene vector to generate transgenic mice Further support for our assumption that the 3.0-kb with the aim of following a developmental profile of promoter fragment is likely to function adequately in reporter gene expression, a strategy employed success- transgenic mice comes from studies with mice trans- fully previously (Goring et al., 1987; Rossant et al., 1991). genic for the CD155 gene. These mice that were gener- To corroborate the pattern of CD155 promoter activity in ated with CD155 upstream sequences of similar length transgenic mice, we performed immunohistological anal- develop neurological symptoms identical to primate po- yses for the detection of CD155 in human embryonic liomyelitis when injected with poliovirus (Ren et al., 1990; tissues. Koike et al., 1991; Gromeier et al., 1996). This indicates In CD155/LacZ tg mice, the expression of lacZ under that the function of the CD155 poliovirus receptor in the control of upstream sequences of CD155 was de- CD155 tg mice mimics that observed in primates. Based tected in anterior midline structures of the developing on these results, we believe that the CD155 promoter CNS during a defined period of embryonic development. fragment employed in our studies harbors the necessary Transgene expression was highly restricted to floor plate cis-acting elements to generate cell type-specific gene and notochord. These structures determine the ventral expression in vivo. polarity of the spinal cord and differentiation of motor Genotypical analysis of newborn CD155/LacZ tg mice neurons (Tanabe and Jessell, 1996; Placzek et al., 1993). generated from germline manipulated embryos revealed Interestingly, CD155 promoter activity was also detected the presence of the transgene in four independent in the nerve fiber layer of the retina, the optic nerve, and founder lines. All four independent founder lines were the region of the future optic chiasm. In accordance with examined for the expression pattern of LacZ under the our transgenic studies, CD155 expression was detected control of CD155-specific upstream sequences during in the neuroretina of human embryos in parallel stages of embryonic development. To this end, whole-mount stain- ontogeny. ing of transgenic mouse embryos at various gestational CD155 promoter activity within the evolving embryonic stages with the LacZ substrate X-gal (see Materials and CNS is restricted to structures giving rise to spinal an- Methods) was performed (Fig. 3). At the gestational age terior horn motor neurons, a population of neural cells shown [embryo day 12.0 (E12.0)], reporter gene expres- that is uniquely susceptible to poliovirus. This associa- sion in all founder lines was detected in anterior midline tion suggests that the distribution of CD155 may indeed structures spanning the neuraxis from the caudal dien- direct poliovirus tropism toward a select group of spe- cephalon to the lower sacral spinal cord. Although the cialized neurons within the primate CNS. intensity of the signal was delicate in lower lumbar and sacral regions of the spinal cord, it was reproducibly RESULTS seen in all four independent founder lines (Figs. 3 and 4). Generation of CD155/LacZ tg mice Reporter gene expression in the developing brain To generate transgenic mice for the analysis of CD155 promoter activity during embryogenesis, we have con- In E12.0 embryos, CD155 promoter activity specified structed an expression vector containing a 3.0-kb frag- ventral midline structures of the spinal cord that ex- ment of the CD155 upstream region (Solecki et al., 1997, tended cranially beyond the myelencephalon and mes- 1999) linked to the LacZ reporter (Fig. 1B; see Materials encephalon into the caudal diencephalon (Figs. 4A–4G). and Methods). The vector, originally supplied to us by J. Cerebral structures of CD155/LacZ tg mice that ex- 250 GROMEIER ET AL.

FIG. 1. Structure of CD155␣ (A), the CD155/LacZ transgene (B), and the CD155 core promoter (C). (A) The overall domain arrangement of CD155␣ is indicated on top. The lengths of individual domains subdivided by the cysteine disulfide bonds characteristic for the IG-like domains are indicated by numbers. (B) The CD155/LacZ transgene consists of a 3000-bp CD155 upstream sequence (harboring the core promoter), a Shine-Dalgarno/Kozak consensus signal (SDK), and the open reading frame of ␤-galactosidase. (C) The CD155 core promoter (numbering: A of ATG ϭϩ1). The brackets indicate the region containing multiple transcriptional initiation sites (Solecki et al., 1997). The transcribed region is indicated by an open box; a stippled box represents the coding region of exon 1. The location of DNase I footprints known to mediate binding to transcriptional regulators NRF-1 and AP-2 is shown (Solecki et al., 1997, 1999). pressed the reporter are linked to floor plate markers future optic nerve and chiasm as well as the neuroretina common to corresponding regions in the spinal cord at E13.0 was altogether absent in E19.0 embryos (data (Colamarino and Tessier-Lavigne, 1995; Placzek et al., not shown). Nonspecific enzymatic activity was ex- 1993). Maximal levels of LacZ activity were detected pressed within the endolymphatic sac (Fig. 4C), which around E12.0 within the presumptive supraspinal floor stained positive in nontransgenic embryos at all stages plate. At that stage, anterior parts of the mesencephalon (not shown for nontransgenic embryos). lining the median sulcus were delineated by LacZ activity (Figs. 4A and 4B) while the rostral boundary of floor plate Reporter gene expression in the spinal cord during according to LacZ expression was the caudal dienceph- embryonic development alon (Fig. 4A). Focal centers of transgene expression within the anterior midbrain extended caudally through- In CD155/LacZ tg mice, transgene expression was out the myelencephalon delineating the medullary raphe readily detectable within the spinal cord. Measured by (Fig. 4D). In addition to regions believed to functionally whole-mount staining of transgenic offspring, expression correspond to floor plate, transgene expression was ob- of the reporter gene within the spinal cord became ap- served in other midline structures: developing pituitary parent at approximately E10.5 (data not shown). As with (Fig. 4C) and outlining the path of the optic nerve and staining observed within the developing brain, maximal future region of the optic chiasm (Figs. 4C and 4D). In intensity and most extensive distribution of staining prod- parallel with our findings of CD155 promoter activity in uct throughout the embryonic CNS were observed at the spinal cord (see below) enzymatic activity decreased E12.0 to E12.5 (Figs. 4A–4G). At that time, blue staining rapidly after E12.0. Weak staining of the anterior mesen- product could be distinguished forming an intracranial cephalon, the medullary raphe, and the region of the formation that extended caudally along the spinal axis to DEVELOPMENTAL EXPRESSION OF CD155 IN THE CNS 251

FIG. 2. Reporter gene expression driven by the 3.0-kb CD155 promoter fragment in Ntera-2 and Raji cells after transfection of CD155/ LacZ cDNA. CD155/LacZ cDNA and its promoter-less derivative SD- K.LacZpA128 were transfected into Ntera-2 and Raji cells. The efficiency of all transfections was monitored by cotransfection of experimental plasmids with pRL-TK (a Renilla luciferase expression vector; see Materials and Methods). LacZ activity measured was nor- malized according to the level of Renilla luciferase activity in the transfected cell lysates. Background LacZ activity produced by the promoter-less SDK.LacZpA128 was given the arbitrary value of 1 and the reporter gene activity yielded by transfection of CD155/LacZ was plotted relative to that value. The results shown represent data from three independent experiments.

reach approximately the region corresponding to the FIG. 4. Reporter gene expression within the developing embryonic 42nd somites. CNS. As indicated in the center, sections were obtained from upper LacZ activity was restricted to ventral midline struc- cephalic (A, B), midcephalic (C, D), cervical (C1–C3; E), midthoracic tures of the developing CNS (Figs. 4A–4G). Structures (T3–T6; F), and lumbar regions (L2–L5; G). Serial transcephalic hori- zontal sections of an E12.0 CD155/LacZ tg embryo (A–D) stained with distinguished by reporter gene expression at the level of X-Gal demonstrated the anatomical distribution of CD155 promoter the spinal cord were notochord, floor plate, and the activity. Transgene expression was detected in anterior midline struc- anterior horn proper (Figs. 4E and 4F, compare to Fig. tures from the most rostral part in the anterior pole of the caudal 7D). The set of structures characterized by CD155 pro- diencephalon (A; di) throughout the anterior mesencephalon (A–C; me) moter activity is the source for inductive signals for motor extending throughout the myelencephalon (D; mr) into the high cervical spinal cord (E). Within the spinal cord the main structures visible neuron differentiation and determines ventral spinal cord through positive staining are the floor plate and notochord (E, F; fp, nc). Diffuse staining within the anterior horns of the spinal cord indicates the presence of reporter gene expression in this area as well (E, F). di, diencephalon; pca, posterior cerebral artery; me, mesencephalon; mca, middle cerebral artery; p, pituitary; es, endolymphatic sac; oc, region of future optic chiasm; ba, basilar artery; mr, medullary raphe; fp, floor plate; nc, notochord; ah, anterior horn. Whole-mount stained embryos were cryosectioned at a thickness of 40 ␮m and counterstained with eosin. Scale bars for A–D, E–G, and the center panel represent 0.45 mm.

polarity. A schematic diagram of the anatomical relationship of notochord, floor plate, and ventral spinal cord in the developing CNS is provided in Fig. 7. At E12, the spinal distribution of transgene expression FIG. 3. Whole-mount staining of CD155/LacZ tg embryos with X-Gal. was preponderantly cervico-thoracic with less intense (A) An E12.0 embryo shows distribution of blue staining product along the spinal axis extending cranially into the developing brain. (B) Non- staining evident in lumbo-sacral regions of the cord transgenic siblings treated in parallel did not reveal macroscopically (Figs. 4E–4G). We observed the focus of intensity of visible levels of LacZ expression. transgene expression to shift caudally toward the lower 252 GROMEIER ET AL. thoracic and lumbo-sacral areas of the spinal cord at lyzed CD155 expression in human embryos of different E12.5–E13.0 (data not shown). At developmental stages gestational ages (Figs. 6A–6I). Immunohistochemical later than E12.0 intensity of transgene expression dimin- probing for CD155 in the embryonic human retina was ished as judged by the weakening reporter gene activity performed using the monoclonal antibody D171 against in those structures identified to harbor CD155 promoter CD155 (Bernhardt et al., 1994a). This antibody was used activity. Transgenic embryos assayed in perinatal stages to isolate and clone the CD155 gene (Mendelsohn et al., of development (E16.5 to E19) were frozen and cryosec- 1989) and has been tested for specificity in tissue culture tioned prior to the staining reaction. Transverse sections (Bernhardt et al., 1994a,b), in tissues of CD155 tg mice, through the spinal cord of E16.5–E19 embryos thus as well as in postmortem tissues derived from the human treated did not reveal specific LacZ activity (data not CNS. Immunohistochemical experiments using cell lines shown). expressing CD155 as well as tissues derived from CD155 Throughout the short period of histochemically detect- tg mice were performed to optimize staining conditions able CD155 promoter activity in CD155/LacZ tg mouse (data not shown). embryos (approximately E10.5–E13.0) the distribution of Treatment of tissue sections from human ocular tis- transgene expression was limited to anterior midline sues revealed an expression pattern paralleling CD155 structures of the CNS with gradual progression caudally. promoter activity in CD155/LacZ tg mice (Fig. 6; compare CD155 promoter activity extended cranially into areas with Fig. 5). Expression of CD155 was detected within the that have been demonstrated to functionally correspond nerve fiber layer of the embryonic retina (Fig. 6B). As to the floor plate at the level of the spinal cord (see seen in CD155/lacZ tg mice, CD155 promoter-driven ex- below). All four transgenic founder lines analyzed pression may be limited to a midgestational period, be- showed an identical pattern of reporter gene activation, cause expression in human embryos was almost unde- both in anatomical and in temporal distribution. tectable at 16 weeks gestational age (Fig. 6E) and was altogether absent in adult retina (Fig. 6H). CD155 could CD155 promoter activity within the developing ocular not be demonstrated immunohistochemically with cer- system tainty in sections of human embryonic spinal cord (data not shown). This may be due to the locally and tempo- In addition to ventral midline structures along a major rally restricted expression of CD155 and the difficulty of extension of the neuraxis we detected CD155 promoter obtaining human embryonic tissues of those defined activity in the developing visual system (Figs. 5A–5C). areas exhibiting CD155 promoter activity at a certain Expression of the reporter gene was observed in ret- gestational age. Furthermore, as observed in CD155/ inofugal fibers constituting the nerve fiber layer of the LacZ tg mice, in which CD155 promoter activity was developing retina. Positively staining axons converged in clearly stronger in the neuroretina, expression within the the optic disk and extended through the optic stalk along developing human spinal cord may be weaker than in the the path of progression of the optic nerve toward the ocular system (Figs. 3, 4, and 5). region of the future optic chiasm (Figs. 4C, 5B, and 5C). CD155 promoter activity within retinal ganglion cells pro- gressed in parallel with axonogenesis and did not in- DISCUSSION volve the retinal ganglion cell somata (Fig. 5C). The The isotypes of the cell surface proteins now recog- temporal pattern of CD155 promoter activity during on- nized as CD155 have been uncovered because they togeny of the visual system overlapped with the peak of function as the human receptor for poliovirus (Mendel- reporter gene expression within the spinal cord and sohn et al., 1989; Koike et al., 1990). The physiological brain. At E12.5, staining of axonal bundles within the function(s) of these proteins, however, has thus far nerve fiber layer (Figs. 5B and 5C) extended along the eluded identification. This is partly due to the exceed- path of optic nerve advance into the optic stalk (for ingly low levels of CD155 expression in postpartal human anatomical orientation refer to Fig. 5A) with some LacZ tissues precluding conclusive immunohistological detec- activity within the region of the future optic chiasm (Fig. tion of CD155 (Guo and Wimmer, unpublished results). 4C). Reporter gene activity was detected simultaneously The expression pattern and function of CD155, however, within the developing lens (Figs. 5B and 4C). We did not are likely to shed light also on the unique mechanism of detect reporter gene expression within the optic tract the pathogenesis of poliomyelitis. Neuropathological beyond the region of the future optic chiasm toward the damage due to poliovirus infection is strikingly specific, tectum. affecting only a restricted number of spinal cord motor neurons (Wimmer et al., 1993; Gromeier et al., 1995) to Expression of CD155 during primate embryonic give rise to the characteristic clinical features of paralytic development poliomyelitis. While strong evidence exists for cell inter- To corroborate the activity pattern of the CD155 pro- nal mechanisms codetermining cell type specificity of moter during midgestation in transgenic mice, we ana- poliovirus (Gromeier et al., 1996), the influence of distri- DEVELOPMENTAL EXPRESSION OF CD155 IN THE CNS 253 bution of CD155 on the neuropathogenic profile of polio- GRASP (Pollerberg and Mack, 1994), or L1 (Bartsch et al., virus remained unclear. 1989). A role in the mediation of cell adhesion has been CD155 has revealed structural similarities with Ig-like demonstrated for several members of the CD155-related glycoproteins of insects that are expressed during de- family of genes [e.g., hPRR2 (Lopez et al., 1998), hPRR1 velopment of the nervous systems (Hellen et al., 1989; (Takahashi et al., 1999), and mPRR2 (Aoki et al., 1997)]. Seeger et al., 1988; Karlstrom et al., 1993; Ramos et al., These findings and the observed developmental expres- 1993). Other Ig-like cell-surface glycoproteins have also sion pattern of CD155 that matches known adhesion been shown to play a crucial role in the development of molecules of the IGSF suggest a physiological function the CNS in mammals (see below). Therefore, we consid- involving the mediation of cell adhesion. ered it possible that CD155 expression may be develop- The cell-type- and tissue-specific activity of the 3.0-kb mentally regulated during embryogenesis of the nervous CD155 upstream region in transgenic embryos indicated system. that this genomic fragment harbors cis-acting elements The strategy to gain insight into the expression pat- specific for directing gene expression to a select subset terns and possible function(s) of CD155 was based on of structures within the developing CNS. Genetic analy- studies of the CD155 promoter (Solecki et al., 1997, 1999, sis of the CD155 promoter in tissue culture identified a 2000) that have allowed us to generate CD155/LacZ tg core fragment that possesses four cis-acting elements. mice. The analyses described here suggest that the These motifs together with their corresponding cellular CD155 promoter in transgenic animals is selectively ac- transcription factors are likely to contribute to the expres- tive in ventral midline structures of the embryonic CNS sion profile observed in vivo (Solecki et al., 1997, 1999, that determine spinal polarity and spinal cord motor 2000). neuron induction. CD155 promoter activity became ap- We have shown that members of the activator pro- parent around E10.5 and was observed in the notochord tein-2 (AP-2) family of transcription factors bind two ad- and floor plate extending cranially throughout the mye- jacent cis-elements within the CD155 core promoter frag- lencephalon and mesencephalon into the caudal dien- ment (Fig. 1; Solecki et al., 1999). Reporter gene expres- cephalon, matching the previously drawn borders of floor sion driven by the AP-2-responsive CD155 promoter plate extension (Colamarino and Tessier-Lavigne, 1995; matches the distribution of AP-2 transcription factors in Placzek et al., 1993; Ruiz y Altaba et al., 1993). CD155 the developing neural tube (Mitchell et al., 1991) as well promoter activity was also observed coincidental with as the developing retina (Bisgrove et al., 1997). In addi- retinal ganglion cell axonogenesis and extended along tion to AP-2, we have identified nuclear respiratory fac- the path of the optic nerve into the region of the future tor-1 (NRF-1) to be required for CD155 promoter function optic chiasm. (Fig. 1; Solecki et al., 2000). Like AP-2, NRF-1 belongs to The sequence of maturation events leading to the a family of developmentally regulated transcription fac- differentiation of motor neurons in the spinal cord (the tors (Becker et al., 1998; Virbasius et al., 1993). The nrf sole target of poliovirus in the primate CNS) is schemat- gene, which encodes a protein that shares over 90% ically represented in Fig. 7. Short-range and long-range amino acid identity with human NRF-1, is expressed inductive signals stemming from notochord and floor throughout the developing ocular system and CNS of plate stimulate floor plate and motor neuron differentia- zebrafish, paralleling the expression profile of CD155 tion, respectively (reviewed in Tanabe and Jessell, 1996). displayed in CD155/LacZ tg mice (Becker et al., 1998). These stimulating signals are transmitted through sonic The regulation of CD155 promoter activity by transcrip- hedgehog, a factor secreted by notochord and floor plate tion factors that display their function during embryonic cells that has been shown to be essential for motor development of the CNS provides strong corroborating neuron induction. Strong support for the described as- evidence supporting the in vivo promoter activity we have sociation of CD155 expression with the developing ven- observed. tral spinal cord stems from observations of CD155 pro- While CD155 can be readily detected immunohisto- moter activation by sonic hedgehog (Solecki et al., un- chemically in the human embryonic retina, attempts to published results). demonstrate CD155 expression in the adult human or The temporospatial pattern of CD155 promoter activity transgenic mouse CNS have failed. In previous studies, in CD155/LacZ tg mice was confirmed to correlate to Northern blot analyses revealed significant levels of CD155 expression during human development by analy- CD155-related mRNA in many organs of CD155 trans- ses of embryonic tissues of human origin. Expression of genic mice, including those that do not support poliovi- CD155 in the human embryonic neuroretina was re- rus replication (Ren et al., 1990; Koike et al., 1991). These stricted to the nerve fiber layer around midgestation with experiments should be considered with caution since rapidly declining levels toward partum. they employed crude tissue extracts that contain con- The expression pattern displayed by CD155 is similar taminating blood cell populations known to express to known cell adhesion molecules belonging to the IGSF, CD155 (Freistadt et al., 1993). e.g., Nr-CAM/Bravo (Grumet, 1997), SC1/BEN/DM- Our observations indicate that, similar to its insect 254 GROMEIER ET AL.

FIG. 5. Reporter gene expression within the eye of CD155/LacZ tg mice. (A) The embryonic eye at E12.5 (reproduced with permission from Kaufman). pe, pigment epithelium; nr, neural retina; od, future optic disk; l, lens; os, optic stalk. (B) The embryonic eye of CD155/LacZ tg mice at E12.5 (180-␮m cryosection). Reporter gene staining product is present within the innermost aspect of the neural retina: retinofugal axons emanating from retinal ganglion cells to form the nerve fiber layer (nfl) extend through the future optic disk into extraocular tissue from the optic stalk. (C) Detail of B; staining product is distributed along the retinofugal axons (arrow) but is absent from retinal ganglion cell somata. Staining of retinofugal fibers extending through the optic disk into the optic stalk indicates reporter gene expression in the developing optic nerve. Scale bars represent 200 ␮m. FIG. 6. Immunohistological detection of CD155 in embryonic and adult human retina. Cryosections from fresh frozen embryonic [10 weeks postconception (pc)] (A, B), 16 weeks pc (D, E), and adult (G, H) ocular tissues were subjected to immunohistology as described under Materials and Methods. For orientation, age-matched cross sections of the developing human retina indicating the anatomical position of the nerve fiber layer and DEVELOPMENTAL EXPRESSION OF CD155 IN THE CNS 255

FIG. 7. Schematic diagram of the anatomical relationship of structures inducing spinal cord motor neuron differentiation. Midline folding of the neural plate (A) produces the neural fold (B). Secretion of sonic hedgehog (indicated by arrows and arrowheads) by notochord (nc) cells induces floor plate (fp) formation in midline neural plate cells. Closure of the neural tube (C) defines the ventral and dorsal commissures. Motor neurons (mn) differentiate within the ventral spinal cord anterior horn (ah; D) in close proximity to the floor plate. Note the distribution of CD155 promoter activity to match elements implied in motor neuron differentiation (compare with Fig. 4E–4F). Modified from Tanabe and Jessell (1996).

IGSF relatives, CD155 expression may be developmen- Toronto, Ontario, Canada). The reporter gene construct tally regulated with levels peaking during midgestation. was amplified, isolated, and purified by cesium chloride This would explain the difficulty in detecting CD155 gradient centrifugation, run twice through a Sephadex polypeptides in the adult human or CD155 transgenic G50 column (Pharmacia), and cut with restriction endo- mouse CNS. Levels of CD155 expression in the postpar- nucleases BsaAI and BamHI to remove vectorial se- tal CNS may simply be too low to be detected immuno- quences. histochemically. Since susceptibility to poliovirus per- sists throughout life, we assume that decreasing levels Cell culture, transfection, and reporter gene assays of CD155 expression in the postpartal spinal cord are sufficient to mediate infectivity of poliovirus. Ntera-2 (human teratocarcinoma) cells were grown in We have provided evidence for selective CD155 pro- Dulbecco’s minimal essential medium and 10% fetal bo- moter activity in structures of the evolving CNS giving vine serum (FBS). Raji cells (human Burkitt’s lymphoma) rise to spinal cord anterior horn motor neurons. Expres- were grown in RPMI 1640 and 10% FBS. Ntera-2 cells sion of CD155 limited to spinal cord anterior horn motor were transfected with CD155/LacZ cDNA or the promot- neurons may be responsible for the exceedingly specific er-less derivative of this vector (SDK.LacZpA128) using neuropathogenic profile of poliovirus and the distinctive the Lipofectamine Plus Kit (Gibco Life Technologies) fol- clinical manifestations associated with poliovirus neuro- lowing the manufacturer’s instructions. Raji cells were invasion. transfected with these constructs via electroporation at 340 V/1180 ␮F with a BRL electroporator (electrode gap MATERIALS AND METHODS 0.4 cm). Cells were allowed to sit for 10 min at RT and were then transferred with fresh medium/serum into cul- Cloning of the reporter gene construct ture flasks. All transfections were standardized by co- A fragment encompassing 3000 nt of the upstream transfection of pRL-TK (Promega) to monitor transfection region of the CD155 gene (Fig. 1; GenBank Accession efficiency by measurement of Renilla luciferase. Trans- No. X94226) was obtained by PCR from a genomic clone fected cells were harvested 18 h posttransfection for the using primers 5Ј ccggatccagaagtttgaaaacc and 3Ј ggct- measurement of ␤-galactosidase or luciferase activity. gcagccagttgctccgagcagctgg. The putative promoter re- Cells were lysed with passive lysis buffer (Promega). gion of the CD155 gene was inserted into the PstI site of ␤-Galactosidase activity was assessed photometrically vector pSDK.LacZpA128 containing the open reading by using the substrate chlorophenol red-␤-D-galactopyr- frame of ␤-galactosidase (kindly provided by J. Rossant, anoside. Luciferase enzymatic activity was measured

the pigment epithelium are shown in C, F, and I. A, D, and G show sections treated with an IgG1 isotype-matched nonspecific control antibody. As expected, no staining was observed when ocular sections were incubated with the nonspecific control antibody. B, E, and H were treated with a monoclonal antibody specifically recognizing CD155 (antibody D171). B shows strong expression of CD155 in the nerve fiber layer (arrow) of the embryonic human retina at 10 weeks pc (compare to expression of LacZ in the nerve fiber layer of CD155/LacZ tg mice; Fig. 5B). At 16 weeks pc (E), CD155 expression was barely detectable in the nerve fiber layer, and the presence of CD155 could no longer be demonstrated in the adult human retina (H). The scale bar represents 250 ␮m. 256 GROMEIER ET AL. using the Dual Luciferase Assay Kit following the man- procedures. Human tissues were obtained as discarded ufacturer’s instructions (Promega). material from the Department of Pathology, Duke Univer- sity Medical Center, using an Institutional Review Board Generation of transgenic mice approved protocol. Gestational age was determined by Transgenic mice were generated in an inbred C57/Bl6 dates and measuring the crown–rump length and foot background by standard procedures (Hogan et al., 1986). size. Transgenic offspring were identified by dot blot hybrid- ization analysis of genomic DNA using a [␣-32P]dCTP- ACKNOWLEDGMENTS labeled LacZ DNA probe. Male transgenic mice were We thank J. Rossant for providing the pSDK.LacZpA128 vector and G. mated with C57/Bl6 females for propagation of individual Bernhardt for help with construction of the transgene. Adult human founder lines. Females mated with transgenic males neural tissue specimens were donations from the Neurological Re- were sacrificed to examine fetuses for LacZ expression search Specimen Bank, VA Greater Los Angeles Healthcare System, in midgestation. West Los Angeles Healthcare Center, Los Angeles, California, which is sponsored by NINDS–NIMH, the National Multiple Sclerosis Society, and the Veterans Health Services and Research Administration, De- 5-Bromo-4-chloro-3-indolyl-␤-D-galactopyranoside partment of Veterans Affairs. We also thank B. F. Haynes for access to (X-Gal) staining of mouse embryos and tissue his repository of human tissues. D.S. was a member of the graduate processing program in Molecular and Cellular Biology, SUNY at Stony Brook. This work was supported by NIH Grant AI39485. M.G. is a recipient of a Female C57/Bl6 mice were mated with male trans- Burroughs–Wellcome Fund career award in the Biomedical Sciences. genic founder animals or the male transgenic offspring of female transgenic founder mice. Timed-pregnant fe- REFERENCES male mice were sacrificed and perfused with phosphate- buffered saline (PBS), and their embryos were removed. Aoki, J., Koike, S., Asou, H., Isa, E., Suwa, H., Tanaka, T., Miyasaka, M., Embryos were fixed in 2% neutral buffered paraformal- and Nomoto, A. (1997). Mouse homolog of poliovirus receptor-related dehyde/0.2% glutaraldehyde for2hatroom temperature. gene 2 product, mPRR2, mediates homophilic cell aggregation. Exp. Cell Res. 15, 374–384. After being rinsed for 2 h with PBS at 4°C, embryos were Bartsch, U., Kirchhoff, F., and Schachner, M. (1989). Immunohistological submerged in the staining solution containing 5 mM localization of the adhesion molecules L1, N-CAM, and MAG in the K3[Fe(CN)6]and5mMK4[Fe(CN)6], 2 mM MgCl, 0.1% developing and adult optic nerve of mice. J. Comp. Neurol. 284, deoxycholate/NP-40, and 1 mg/ml X-Gal (Sigma) in PBS. 451–462. Staining was allowed to proceed for 36 h at 36°C, and Becker, T. S., Burgess, S. M., Amsterdam, A. H., Allende, M. L., and Hopkins, N. (1998). not really finished is crucial for development of embryos were rinsed for4hinPBSandpostfixed for 30 the zebrafish outer retina and encodes a transcription factor highly min in 10% neutral buffered paraformaldehyde at 4°C. homologous to human nuclear respiratory factor-1 and avian initia- After embryos were rinsed for2hinPBSat4°Cthey tion binding repressor. Development 125, 4369–4378. were submerged in 30% sucrose in PBS overnight, Belnap, D. M., McDermott, B. M., Filman, D. J., Cheng, N., Trus, B. L., placed in OCT compound, and gently frozen at Ϫ80°C for Zuccola, H. J., Racaniello, V. R., Hogle, J. M., and Steven, A. C. (2000). Three-dimensional structure of poliovirus receptor bound to poliovi- 20 min for cryostat sectioning. Cryostat sections were rus. Proc. Natl. Acad. Sci. USA 97, 73–78. obtained at a thickness of 40 or 180 ␮m, placed on Bernhardt, G., Bibb, J. A., Bradley, J., and Wimmer, E. (1994a). Molecular silane-treated microscope slides (see below), dried at characterization of the cellular receptor for poliovirus. Virology 199, room temperature, and counterstained with eosin. 105–113. Bernhardt, G., Harber, J. J., Zibert, A., deCrombrugghe, M., and Wimmer, Immunohistochemistry E. (1994b). The poliovirus receptor: Identification of domains and amino acid residues critical for virus binding. Virology 203, 344–356. Cryosections from human tissues were prepared at a Bisgrove, D. A., Monckton, E. A., and Godbout, R. (1997). Involvement of thickness of 4 ␮m, fixed in cold acetone, and rinsed with AP-2 in regulation of the R-FABP gene in the developing chick retina. Mol. Cell. Biol. 17, 5935–5945. PBS. Sections were blocked with PBS containing 1% Chadeneau, C., LeCabellec, M., LeMoullac, B., Meflah, K., and Denis, bovine serum albumin and incubated with 5.5 ng/ml M. G. (1994). Over-expression of a novel member of the immunoglob- affinity-purified anti-CD155 monoclonal antibody or an ulin superfamily in Min mouse intestinal adenomas. Int. J. Cancer 68, IgG1 isotype-matched control antibody for1hatroom 817–821. temperature. After three washes with PBS, the sections Colamarino, S. A., and Tessier-Lavigne, M. (1995). The role of the floor plate in axon guidance. Annu. Rev. Neurosci. 18, 497–529. were incubated with biotinylated anti-mouse polyclonal Eberle, F., Dubreuil, P., Mattei, M. G., Devilard, E., and Lopez, M. (1995). antibody (Vector Laboratories, Burlingame, NH) for 1 h at The human PRR2 gene, related to the human poliovirus receptor room temperature and washed with three exchanges of gene (PVR), is the true homolog of the murine MPH gene. Gene 4, PBS. Sections were then incubated with a horseradish 267–272. peroxidase–streptavidin complex (Vector Laboratories) Freistadt, M., Fleit, H. H., and Wimmer, E. (1993). Poliovirus receptor on human blood cells: A possible extraneural site of poliovirus replica- for 30 min, rinsed with PBS, and stained with di-amino- tion. Virology 195, 798–804. benzidine for 5 min. Sections were counterstained with Geraghty, R. J., Krummenacher, C., Cohen, G. H., Eisenberg, R. J., and hematoxylin, dehydrated, and mounted using standard Spear, P. G. (1998). Entry of alphaherpesviruses mediated by polio- DEVELOPMENTAL EXPRESSION OF CD155 IN THE CNS 257

virus receptor-related protein 1 and poliovirus receptor. Science 280, cells and is involved in mediating their extension on axons. Dev. Biol. 1618–1620. 165, 670–687. Goring, D. R., Rossant, J., Clapoff, S., Breitman, M. L., and Tsui, L. C. Ramos, R. G. P., Igloi, G. L., Lichte, B., Baumann, U., Maier, D., Schnei- (1987). In situ detection of ␤-galactosidase in lenses of transgenic der, T., Brandstaetter, J. H., Froehlich, A., and Fischbach, K. F. (1993). mice with a gamma-crystallin/lacZ gene. Science 235, 456–458. The irregular chiasm C-roughest locus of Drosophila, which affects Gromeier, M., Lu, H. H., and Wimmer, E. (1995). Mouse-neurovirulent axonal projections and programmed cell death, encodes a novel poliovirus strains cause damage in the central nervous system immunoglobulin-like protein. Genes Dev. 7, 2533–2547. distinct from poliomyelitis. Microb. Pathogen 18, 253–267. Ren, R., Costantini, F., Gorgacz, E. J., Lee, J. J., and Racaniello, V. R. Gromeier, M., Alexander, L., and Wimmer, E. (1996). Internal ribosomal (1990). Transgenic mice expressing a human poliovirus receptor: A entry site substitution eliminates neurovirulence in intergeneric po- new model for poliomyelitis. Cell 63, 353–362. liovirus recombinants. Proc. Natl. Acad. Sci. USA 93, 2370–2375. Rossant, J., Zirngibl, R., Cado, D., Shago, M., and Giguere, V. (1991). Grumet, M. (1997). Nr-CAM: A cell adhesion molecule with ligand and Expression of a retinoic acid response element-hsplacZ transgene receptor functions. Cell Tissue Res. 290, 423–428. defines specific domains of transcriptional activity during mouse He, Y., Bowman, V. D., Mueller, S., Bator, C. M., Bella, J., Peng, X., Baker, embryogenesis. Genes Dev. 5, 1333–1344. T. S., Wimmer, E., Kuhn, R. J., and Rossmann, M. G. (2000). Interaction Ruiz y Altaba, A., Prezioso, V. R., Darnell, J. D., and Jessell, T. M. (1993). of the poliovirus receptor with poliovirus. Proc. Natl. Acad. Sci. USA Sequential expression of HNF-3b and HNF-3a by embryonic orga- 97, 79–84. nizing centers: The dorsal lip/node, notochord and floor plate. Mech. Hellen, C. U. T., deCrombrugghe, M., Wimmer, E., Seeger, M. A., and Dev. 44, 91–108. Kaufman, T. C. (1989). Amalgamated sequences. Nature 340, 682. Seeger, M. A., Haffley, L., and Kaufmann, T. C. (1988). Characterization Hogan, B., Constantini, F., and Lacy, E. (1986). “Manipulating the Mouse of amalgam: A member of the immunoglobulin superfamily from Embryo: A Laboratory Manual.” Cold Spring Harbor Laboratory Drosophila. Cell 55, 589–600. Press, Plainview, NY. Solecki, D., Schwarz, S., Wimmer, E., Lipp, M., and Bernhardt, G. (1997). Karlstrom, R. O., Wilder, L. P., and Bastiani, M. J. (1993). Lachesin: An The promoters from human and monkey poliovirus receptors. J. Biol. immunoglobulin superfamily protein whose expression correlates Chem. 272, 5579–5586. with neurogenesis in grasshopper embryos. Development 118, 509– Solecki, D., Wimmer, E., Lipp, M., and Bernhardt, G. (1999). Identification 522. and characterization of the cis-acting elements of the human CD155 Kaufman, M. H. (1992). “Atlas of Mouse Development.” Academic Press, gene core promoter. J. Biol. Chem. 274, 1791–1800. London, UK. Solecki, D., Bernhardt, G., Lipp, M., and Wimmer, E. (2000). Identification Koike, S., Horie, H., Ise, I., Okitsu, A., Yoshida, M., Iizuka, N., Takeuchi, of a nuclear respiratory factor-1 binding site within the core promoter K., Takegami, T., and Nomoto, A. (1990). The human poliovirus re- of the human poliovirus receptor/CD155 gene. J. Biol. Chem. 275, ceptor protein is produced both as membrane-bound and secreted 12453–12462. forms. EMBO J. 9, 3217–3229. Takahashi, K., Nakanishi, H., Miyahara, M., Mandai, K., Satoh, K., Satoh, Koike, S., Taya, S., Kurata, T., Abe, S., Ise, I., Yonekawa, H., and Nomoto, A., Nishioka, H., Aoki, J., Nomoto, A., Mizoguchi, A., and Takai, Y. A. (1991). Transgenic mice susceptible to poliovirus. Proc. Natl. Acad. (1999). Nectin/PRR: An immunoglobulin-like cell adhesion molecule Sci. USA 88, 951–955. recruited to cadherin-based adherens junctions through interaction Koike, S., Ise, I., Sato, Y., Yonekawa, H., Gotoh, O., and Nomoto, A. with Afadin, aPCZ domain-containing protein. J. Cell Biol. 145, 539– (1992). A second gene for the African green monkey poliovirus receptor that has no putative NM-glycosylation site in the functional 549. N-terminal immunoglobulin-like domain. J. Virol. 66, 7059–7066. Tanabe, Y., and Jessell, T. M. (1996). Diversity and pattern in the Lopez, M., Eberle, F., Mattei, M. G., Gaert, J., Birg, G., Bardin, F., Maroc, developing spinal cord. Science 274, 1115–1123. C., and Dubreuil, P. (1995). Complementary DNA characterization and Tessier-Lavigne, M., and Goodman, C. S. (1996). The molecular biology chromosomal localization of a human gene relate to the poliovirus of axon guidance. Science 274, 1123–1133. receptor-encoding gene. Gene 3, 261–265. Virbasius, C. A., Virbasius, J. V., and Scarpulla, R. C. (1993). NRF-1, an Lopez, M., Aoubala, M., Jordier, F., Isnardon, D., Gomez, S., and activator involved in nuclear–mitochondrial interactions, utilizes a Dubreuil, P. (1998). The human poliovirus receptor related 2 protein is new DNA-binding domain conserved in a family of developmental a new hematopoietic/endothelial homophilic adhesion molecule. regulators. Genes Dev. 7, 2431–2445. Blood 92, 4602–4611. Warner, M. S., Geraghty, R. J., Martinez, W. M., Montgomery, R. I., Mendelsohn, C. L., Wimmer, E., and Racaniello, V. R. (1989). Cellular Whitbeck, J. C., Xu, R., Eisenberg, R. J., Cohen, G. H., and Spear, P. G. receptor for poliovirus: Molecular cloning, nucleotide sequence, and (1998). A cell surface protein with herpesvirus entry activity (HveB) expression of a new member of the immunoglobulin superfamily. Cell confers susceptibility to infection by mutants of herpes simplex virus 56, 855–865. type 1, herpes simplex virus type 2, and pseudorabies virus. Virology Mitchell, P. J., Timmons, P. M., Hebert, J. M., Rigby, P. W., and Tjian, R. 246, 179–189. (1991). Transcription factor AP-2 is expressed in neural crest cell Wimmer, E., Harber, J. J., Bibb, J. A., Gromeier, M., Lu, H. H., and lineages during mouse embryogenesis. Genes Dev. 5, 105–119. Bernhardt, G. (1993). Poliovirus receptors. In “Cellular Receptors for Morrison, M. E., and Racaniello, V. R. (1992). Molecular cloning and Animal Viruses” (E. Wimmer, Ed.), pp. 101–128. Cold Spring Harbor expression of a murine homolog of the human poliovirus receptor Laboratory Press, Plainview, NY. gene. J. Virol. 66, 2807–2813. Yamada, T., Placzek, M., Tanaka, H., Dodd, J., and Jessell, T. M. (1991). Placzek, M., Jessell, T. M., and Dodd, J. (1993). Induction of floor plate Control of cell pattern in the developing nervous system: Polarizing differentiation by contact-dependent, homeogenetic signals. Devel- activity of the floor plate and notochord. Cell 64, 635–647. opment 117, 205–218. Yoshihara, Y., Oka, S., Ikeda, J., and Mori, K. (1991). Immunoglobulin Pollerberg, G. E., and Mack, T. G. (1994). Cell adhesion molecule superfamily molecules in the nervous system. Neurosci. Res. 10, SC1/DMGRASP is expressed on growing axons of retina ganglion 83–105.