toward a genetic analysis of cell-matrix adhesion - biologybiology.hunter.cuny.edu/cellbio/feinstein...

10.1101/SQB.1992.057.01.030Access the most recent version at doi: 1992 57: 249-258Cold Spring Harb Symp Quant Biol

R.O. Hynes, E.L. George, E.N. Georges, et al. Toward a Genetic Analysis of Cell-Matrix Adhesion

References

http://symposium.cshlp.org/content/57/249#related-urlsArticle cited in:

http://symposium.cshlp.org/content/57/249.refs.htmlThis article cites 55 articles, 25 of which can be accessed free at:

serviceEmail alerting

click herethe box at the top right corner of the article orReceive free email alerts when new articles cite this article - sign up in

http://symposium.cshlp.org/subscriptions go to: Cold Spring Harbor Symposia on Quantitative BiologyTo subscribe to

Copyright © 1992 Cold Spring Harbor Laboratory Press

Cold Spring Harbor Laboratory Press on April 29, 2009 - Published by symposium.cshlp.orgDownloaded from

http://symposium.cshlp.org/lookup/doi/10.1101/SQB.1992.057.01.030

http://symposium.cshlp.org/content/57/249.refs.html

http://symposium.cshlp.org/content/57/249#related-urls

http://symposium.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=sqb;57/0/249&return_type=article&return_url=http://symposium.cshlp.org/content/57/249.full.pdf

http://symposium.cshlp.org/subscriptions

http://symposium.cshlp.org/

http://www.cshlpress.com

Toward a Genetic Analysis of Cell-Matrix Adhesion

R.O. HYNES, E.L. GEORGE, E.N. GEORGES,* J . -L. GUAN,t H . RAYBURN, AND J .T. YANG

Howard Hughes Medical Institute and Center for Cancer Research, Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

The interactions of cells with their neighbors and with the extracetlular matrix play crucial roles in development and in numerous physiological and pathological processes (Hay 1991; Hynes and Lander 1992). Studies of cell adhesion during the past decade and more have uncovered the existence of a large number of cell adhesion molecules. These include cell- surface receptors involved in cell-cell adhesion, such as the cadherins (Takeichi 1988, 1990, 1991), members of the immunoglobulin superfamily (Jessell 1988; Grumet 1991), and selectins (Bevilacqua et al. 1991; Lasky and Rosen 1992). Each of these families of receptors includes multiple related molecules. The same is true for cell-matrix adhesion, which is most frequently mediated by the family of adhesion receptors known as integrins (Hynes 1987, 1992; Albelda and Buck 1990; Hemler 1990).

Integrins are heterodimeric transmembrane receptors whose large extracellular domains interact with adhesive molecules of the extracellular matrix, or with counterreceptors on other cells, to mediate cell adhesion. Their smaller cytoplasmic domains interact with the cytoskeleton and are also involved in signal transduction processes (Shattil and Brugge 1991; Hynes 1992). The integrin family is very diverse (Table 1), and although each integrin shows selectivity in the ligands it binds, there is degeneracy of two kinds. First, many integrins bind several different ligands. Second, most extracellular matrix proteins can be recognized by multiple integrins. This is diagramed for two adhesive extracellular matrix molecules (laminin and fibrinogen) in Figure 1. Most cells express multiple integrins and are, therefore, able to interact with extracellular matrix molecules with considerable versatility. Further complexity is added by the fact that several integrin subunits can occur in alternatively spliced forms (see aster- isks in Table 1). In the case of vertebrate integrins, this alternative splicing is only known to affect the cytoplasmic domains, where it is thought to modulate interactions with the cytoskeleton and/or signal transduction events. However, in Drosophila, at least two integrin subunits can be alternatively spliced in their extracellular domains, in regions of the molecules close to the ligand-binding and/or subunit interaction sites

Present addresses: *LGME/CNRS, U184 INSERM, Facult6 de M6dicine, 67000 Strasbourg, France; tDepartment of Pathology, Cor- nell University School of Veterinary Medicine, Ithaca, New York 14853.

(Brown et al. 1989; G. Yee et al., in prep.). This raises the possibility that the alternative splicing may alter ligand-binding affinity or specificity.

Whereas some extracellular matrix proteins such as collagens and laminins are encoded by multigene families, alternative splicing also introduces important variations in several extracellular matrix molecules. Both fibronectin (FN) (Hynes 1990) and tenascin (Spring et al. 1989; Gulcher et al. 1991; Siri et al, 1991) are alternatively spliced to produce multiple isoforms. In the case of FN, one of the alternatively spliced segments includes a site for interaction with a specific integrin (Wayner et al. 1989; Guan and Hynes 1990; Mould et al. 1990, 1991). The alternative splicing of FN and the interaction of two different integrins with different sites in FN are diagramed in Figure 2.

Given this complexity of both integrins and their ligands, it is a challenging problem to decide which integrin-matrix interactions are involved and crucial in any given biological process. The most common approach has been to use blocking antibodies or peptides

Table 1. The Integrin Receptor Family

Subunits Ligands and counterreceptors

8"

% collagens/LM a 2 collagens / LM a* .~ FN/LM/collagens a 4 FN(V25)/V-CAM-1 % FN(RGD) oe* LM a v LM a s ? a v V N / F N (?)

a L 1-CAM- 1 / I -CAM-2 /32 a M i C 3 b / F B / F X / I - C A M - 1

a x FB/ iC3b?

J ~ OL|I b F B / F N / V W F / V N / T S P av VN / FB / V W F /TSP/ FN / OP/ colIagen

/3" a~ LM?? 13 5 a v VN 136 a v FN /37(=/3p) a 4 FN(V25)/V-CAM-l? /38 av ?

Abbreviations: (C3bi) C3b component of complement, inactivated; (FB) fibrinogen; (FN) fibronectin; (FX) factor X; (I-CAM) inter- cellular adhesion molecule; (LM) laminin; (OP) osteopontin; (TSP) thrombospondin; (V-CAM-l) vascular cell adhesion molecule; (VN) vitronectin; (VWF) yon Willebrand factor.

*Alternatively spliced.

Cold Spring Harbor Symposia on Quantitative Biology, Volume LVII. �9 1992 Cold Spring Harbor Laboratory Press 0-87969-063-1/92 $3.00 249




250 HYNES ET AL.

ctllb[33 ctllb[~3 a. ct VI33 ~v[33 b.

AC~ chain site AC~ chain site ~l ? \ / \ J

\~ RGDS / \~ RGDS// " " "

\,/'

~ M ~ 2 ~ ~ ~ ctM~2

/ '~ chain sit~, x c~X[32 .~ ' chain sit~,x

~IHLGCAKQAGD~' IHLGGAKQAGDV l lllllll o~6~l ~I(b[~3 ttllb~3

Figure 1. Multiple integrins binding to different parts of adhesive extracellular matrix glycopro- teins. (a) Fibrinogen is recognized by four different integrins; both av~ 3 and a~ib/33 recognize an RGD site in the Aa chain; a]]b/33 in addition recognizes a different site at the carboxyl ter- minus of the 3' chain; aMfl2 binds to fragment D and a~/32 to the amino-terminal disulfide knot (DSK) region. (b) Laminin is recognized by two groups of integrins; alfi I (and probably the homologous aEfli) binds to the cross region, whereas a3/31 and a6/31 (and probably a713 l, which is closely related) bind near the end of the long arm. a6fl ~ may or may not be an additional laminin receptor.

specific for particular integrins and /o r matrix proteins. Al though this approach has provided much informa- t ion, it has its limits, especially for analyses of events in intact organisms. A complementary approach is to use genetic analyses to ablate specific components and dis- sect their functions in vivo. Because of its tractable genetics, Drosophila has been a primary organism for this approach to integrin function. Several groups have analyzed the effects of mutat ions in integrins on various developmenta l processes (Brower and Jaffe 1989; Lep- tin et al. 1989; Wilcox et al. 1989; Wilcox 1990; Zusman

et el. 1992), and we have recently used this system to investigate the functional consequences of alternative splicing (Zusman et al. 1992) and of mutat ions in the cytoplasmic domain (Grinbla t et al. 1992).

It would be advantageous to be able to apply genetic analyses to study the roles of integrin-matrix interactions in vertebrates, and we have initiated a program of research directed toward this aim. We have concentrated initially on FN, its alternatively spliced isoforms, and the subset of integrins that interacts with FN. We summarize here our progress to date.

Fibrin Gelatin Cell Cell Heparin Collagen Heparin Fibrin

a . ) H i ) - . . . . . ~ i ~.~ , , SH SH

,~ ~ ~ i ss EIIIB EIIIA

" " w -

Type II

Type III

v II

V120

V95

V89

VS4

vo

b . EIIIB EIIIA V Heparin Collagen _ . / , Cell . � 9 Heparin,. T , . . I i

I I t ~ 1 ~ - T I I I F ) I r ' ?" s s NH=~ , I I COOH

\ / . . . .

IMATRIX ASSEMBLY~ t S~tERGY----R~D ] I'HEPARIN IIEILOVI( REOV]

INTEGRIN RECEPTORS

Figure 2. A single FN gene can generate up to 20 different FN isoforms. (a) These arise by alternative splicing at three positions designated EIIIB, EIIIA, and V. EIIIB and EIIIA are protein modules (type III repeats), which can be included or excluded by exon skipping. The V region can be fully included or partially or wholly excluded by complex splicing of a single exon; the pattern shown is that observed in humans. (b) Several iutegrins recognize FN (see Table 1). The best characterized are as/3t, which recognizes an RGD sequence present in all isoforms of FN, and a4fl t, which recognizes a different sequence located in the alternatively spliced V25 segment of the V region. Both parts of the figure depict functional binding sites within each FN suhunit.




GENETICS OF CELL-MATRIX A D H E S I O N 251

RESULTS AND DISCUSSION

Distribution of FN Isoforms In Vivo

To provide context for the molecular genetic analyses that follow, we first provide a summary of descriptive studies of the expression of the different FN isoforms in vivo. As shown in Figure 2, there are three regions of alternative splicing of the FN t ranscr ip t - - EII IB, E I I IA , and V (or I l ICS)- -which we refer to as B, A, and V for simplicity. Nucleic acid probes and antibodies specific for each of these segments can be prepared and have been used to define the patterns of expression of the different splice forms in embryos and adult tissues and during physiological and pathological responses. A number of generalizations can be made. First, it appears that early embryos express fibronectins (B § A § V § including all three splice segments (Nor- ton and Hynes 1987; ffrench-Constant and Hynes 1989; DeSimone et al. 1992). This is particularly relevant in situations where extensive cell migration occurs, such as gastrulation, neural crest migration, and develop-

ment of the heart (ffrench-Constant and Hynes 1988). In all these situations, FN is thought to be important in promoting cell migration (Hynes 1990). Second, as development of differentiated cell types proceeds, cer- tain of the alternative segments are omitted in a cell- type-specific fashion (Vartio et at. 1987; Burton-Wurs- ter et at. 1989; ffrench-Constant and Hynes 1989; Glukhova et al. 1989, 1990; Bennett et al. 1991; Pagani et al. 1991); Figure 3 diagrams a few examples. Finally, the pattern of alternative splicing can alter in response to various stimuli. Examples include wound healing (ffrench-Constant et al. 1989), myocardial infarction (Shekhonin et al. 1990; Knowlton et al. 1992), athero- sclerosis (Glukhova et al. 1989, 1990; Mamuya and Brecher 1992), lung injury (Peters et al. 1988), and tumorigenesis (Carnemolla et al. 1989; Oyama et al. 1989). In each case, there appears to be an increase in the inclusion of the B and A segments, which are frequently present at low levels at steady state in adult tissues but increase in response to the various pathological situations. One can view this response as a

a. NEURAL CREST

MIGRATING NEURAL CREST ECTOOERM CELLS FN" ~ / FN++

NEURAL ~ ~ SOMITE TUBE Q FN+ FN§ B+A+V+

B+A+V+

b. CUTANEOUS WOUND HEALING

KERATINOCYTE

c. LIVER d, PECTORAL MUSCLE/STERNUM

SMALL VESSELS B+A+V+

LARGEVESSELS B-A+V§

HEPATOCYTES B-A-V+/-

MUSCLE FIBROBLASTS

B-A+V+

/~ PERICHONDRIUM B+A+V+

N|, '.':::::;:'::::.2: "..--:::..;:--::::-: "::.::::.::::;::.2: "::;:";:::':;2: ":::-:::.:::.:::::-:

Figure 3. Patterns of alternative splicing of FN. Diagrams based on in situ hybridization (and some antibody) detection of the alternatively spliced segments. (a) FN expressed in early embryos and contributing to the matrix in which neural crest cells migrate includes all three splice segments (B § A + V+). This FN is made by several cell types. This pattern of splicing is characteristic of other situations in early embryos where cell migration and proliferation are prominent (e.g., heart, area vasculosa, vasculogenesis). (b) In resting adult skin dermis, the predominant form of FN expressed is B - A-V +. After wounding, the level of FN mRNA rises greatly, and the cells beneath the migrating keratinocytes produce B § A + V + FN. That is~ their pattern of splicing reverts to that characteristic of early embryos (see a). (c) Liver is the site of synthesis of plasma FN (B A V +/-, which is made by hepatocytes, circulates in the blood, and contributes to blood clots (see b), where it can serve as an important adhesive protein. In contrast, endothelial and other cells lining vessels in the liver express FN isoforms, which include the A, and sometimes also the B. segment, (d) The thorax illustrates different patterns of FN splicing by various differentiated cell types; cartilage includes the B segment but not the A; muscle fibroblasts express the inverse pattern, whereas perichondrial fibroblasts include all segments.




252 HYNES ET AL.

reversion to the pattern found in early embryos; this viewpoint highlights the fact that inclusion of the B and A segments correlates with proliferation and/or migration of cells.

Such descriptive analyses of the distribution of FN isoforms can only suggest hypotheses; they cannot test them. To do that, it is necessary to obtain the different isoforms in homogeneous form to test for differential functions or to manipulate the expression of the isoforms.

Molecular Genetic Analysis of FN lsoforms and lntegrins In Vitro

Since all or most tissues express a mixture of different FN isoforms and since cells frequently form disulfide-bonded heterodimers of different isoforms expressed within the same cell (Schwarzbauer et al. 1987, 1989), it is essentially impossible to purify homogeneous FN isoforms from natural sources. Even plasma FN, which lacks the B and A segments, is 50% V § 50% V - (Paul et al. 1986), and V+/V heterodimers are the rule (Schwarzbauer et al. 1989). Since most functional studies on FN have been performed with plasma FN, they have not addressed the functions of the B or A segments. To circumvent this limitation, we have established cell lines expressing full-length FN isoforms encoded by cDNAs (Guan et al. 1990). By introducing the eDNA expression constructs into cells that do not synthesize FN, we were able to produce homogeneous FNs of each defined splicing pattern. Such recombinant-derived FN isoforms can then be used to test for functions dependent on the presence of specific splice segments.

In this way, we were able to demonstrate that lymphoid cells adhere specifically only to those FN forms containing the V segment (Guan and Hynes 1990; Guan et al. 1990). We were also able to show that this adhesion is mediated by the 0/4/31 integrin, which recognizes a short sequence in a part of the V segment (V25) that can be alternatively spliced as a unit (Guan and Hynes 1990). These results conform with data from other laboratories using fragments of plasma FN and synthetic peptides (Wayner et al. 1989; Mould et al. 1990). The lymphoid cells used lack the a5/31 integrin, which recognizes the RGD sequence present in all FNs. However, we introduced an 0/5 cDNA construct into the cells by retroviral transduction and obtained cells expressing 0/5/3t, as well as a4/31. Figure 4a shows that these cells adhere to all forms of FN, irrespective of the presence or absence of the V segment (or any other alternatively spliced segment), in marked contrast with the parent cell line, which expresses 0/4/31 but not a5/31 (Fig. 4b). This experiment demonstrates the use by cells of different FN receptor integrins to discriminate among different FN isoforms. Similar experiments to test for specific functions conferred by the B and A segments are under way but have, so far, failed to reveal a similarly clear distinction.

Molecular Genetic Analysis of FN Isoforms and Integrins In Vivo

To extend these analyses to the in vivo situation and initiate genetic analyses of the roles of different FN isoforms and different integrin receptors for FN, we have turned to molecular genetic approaches in mice. The development of methods for producing transgenic mice (Hanahan 1989) and, more recently, for generat- ing targeted mutations by homologous recombination in embryonic stem (ES) cells (Capecchi 1989) have opened the way to production of mice mutant in genes of interest.

We have used the methods of Capecchi and co- workers (Thomas and Capecchi 1987; Mansour et al. 1988) to generate null mutations for FN and for 0/4 and % integrin subunits. In each case, clones were isolated from mouse genomic libraries that included the 5' end of the gene. Sequencing of each of the clones located the exon encoding the translation initiation site and signal sequence. We then deleted a segment spanning the initiation codon, all or part of the signal sequence, and Y-flanking sequences. In the case of the FN gene, these included the probable site of initiation of transcription as well as parts of the promoter (Patel et al. 1987). It is likely that the same is true for the segments deleted from the 0/4 and 0/5 integrin clones, but these genes have not been analyzed as extensively. The deleted segment in each case was replaced by a cassette comprising the Escherichia coli neomycin-resistance gene driven by the phosphoglycerate kinase (PGK) promoter and followed by a polyadenylation site. Each of the constructs contained 750-850 bp of 5'-flanking sequence from the gene and 5-6 kbp of genomic sequence 3' of the neo r~ insertion (Fig. 5a). In addition, at one or both ends of the construct, we inserted a second cassette comprising the herpes simplex virus thymidine kinase (TK) gene driven by the PGK promoter. The constructs were linearized and introduced into 1 x 108 to 4 x 108 D3 129/Sv ES cells (Doetschman et al. 1985) by electroporation at 25 p~g/ml of DNA. The ES cells were plated for 1 day and then subjected to selection with G418 (200/zg/ml) and gancyclovir (2 p.M) for 6-8 days. The G418 selects for cells expressing the neo R cassette, whereas the gancyclovir selects against cells expressing the TK cassette (Mansour et al. 1988). Doubly resistant cells should include cells in which the neo R cassette has inserted into the targeted gene by homologous recombination (Fig. 5a). Doubly resistant clones were screened for the appropriate re- combinational event by Southern blotting or by PCR and, in each case, we obtained multiple clones with the desired insertional mutation. These were further tested by Southern blotting and grown up for injection into C57BL6 blastocysts (Hogan et al. 1986). Male chimeras with a high level of agouti coat color (derived from the 129/Sv ES ceils) were bred with C57BL6 females and, in each case, we obtained transmission of ES cell-derived gametes. Screening of these agouti progeny by tail blotting revealed transmission of the




a . V- rFNs

GENETICS OF CELL-MATRIX ADHESION

V+ rFNs CONTROLS

pFN

253

INHIBITED by - R G D p e p t i d e s

- anti - 0~5 mAbs

DEPENDENT o n b o t h :-

- 0~4[~1 -- EILDV

- 0~5[~1 -- RGD

r

type IV c o l l a g e n

laminin

b . v- rFNs V+ rFNs CONTROLS

pFN

~ / type IV c o l l a g e n

laminin

INHIBITED by - EILDV p e p t i d e s

- anti - 0~4 mAbs

Figure 4. Cells expressing different integrins recognize different FN isoforms. (a) WEHI 231 cells transfected with a s integrin eDNA express both asfl 1 and a4/31 and bind to all forms of recombinant full-length FN, but not to collagen or laminin. (b) The parent WEH1231 cell line, which expresses a4fl 1 but not a 5 ill, adheres only to FN isoforms including the V segment. This binding can be inhibited by peptides from the V25 segment (containing the sequence EILDV) or by antibodies to a 4 or/31 . In contrast, adhesion of WEHI (as) cells to V isoforms of FN can be inhibited by peptides containing the RGD sequence or by antibodies to a 5 or/31 (panel a).




254 HYNES ET AL.

neo R a. ~ b.

s s

NORMAL I J, -_ -_ NORMAL

LOCUS ~ X T K~ LOCUS

FN-null x B-minus TARGETING ~ ._ TARGETING

VECTOR <s) (s) VECTOR

x I B-plus TARGETED "~ I "" , = TARGETING

LOCUS _ . _ (s) (s~ VECTOR

PROBE

EIIIB

I neoR

\ / \ /

\ / neo R r

TK I I I I

TK l i b

Figure 5. Targeting vectors for homologous recombination of the FN gene. (a) A segment flanked by two restriction sites (S) and containing part of the promoter and first exon is deleted and replaced by a selectable marker, the neomycin-resistance cassette (neoR). DNA containing this replacement plus flanking sequences is fused to a counterselectable marker (TK) and transfected into ES cells. Homologous recombination with the wild-type locus inserts the neo R marker into the genome but deletes the TK marker (see text). Thus, the targeted locus is defective (since it lacks the sites for initiation of transcription and translation) and can be selected on the basis of the neo ~ insertion. Correct recombination can be confirmed by PCR analysis (arrows) and/or by Southern blotting using a probe outside the end (X) of the targeting vector. Exactly analogous strategies were used for the c~ 4 and as genes. (b) The region of the gene encoding the EIIIB segment (in a single exon) is shown. Targeting vectors lacking the EIIIB exon or containing it fused to the adjacent exons were constructed. Each construct also contained the neo R selectable marker in an adjacent intron and the counterselectable marker (TK) at one end. Insertion of these mutations into the gene can be obtained as depicted in a.

mutated alleles with approximately 50% frequency. Therefore, inactivation of the FN gene or of the a 4 or a 5 integrin genes produces recessive mutations, since the heterozygous mutant mice are viable and outwardly normal.

We have therefore been able to establish heterozygous F 1 lines for each of these presumptively null mutations. All three lines have been self-intercrossed. The FN and a 5 lines have been tested by tail blotting and, in both cases, it is clear that the mutations are embryonic lethals; no homozygous mutant progeny have been observed. Progeny of the a 4 line have not been tested as yet. These data clearly demonstrate the importance of FN and a~/3~ integrin for completion of embryonic development.

Analysis of embryonic lethal mutations requires sam- pling of embryos and/or fetuses at different stages of development. The embryos can then be examined, sec- tioned for histological and immunohistochemical analyses, and used to derive cultures for biochemical and cell biological analyses. By the latter procedures, we have confirmed that both the FN and u5 mutations are true null mutations and that the respective proteins are not produced by homozygous mutant embryos. The mor- phological, histological, and immunohistological analyses are most advanced for the FN mutant embryos; some of the results are illustrated briefly here.

FN is first observed in the preimplantation blastocyst (Zetter and Martin 1978; R.O. Hynes et al., unpubl.) but is expressed at higher levels in postimplantation stages. Figure 6a illustrates a day-7.5 embryo at the primitive streak stage stained for FN. FN is concentrated at the interfaces between the three germ layers and is also present within the mesodermal layer, in agreement with published data (Wartiovaara et al.

1979). In contrast, day-7.5 embryos that are homozygous for the FN-null mutation express no detectable FN (Fig. 6b). However, it is of interest to note that the mutant embryo does contain three germ layers. That is, mesoderm forms in the absence of FN expression. This result contrasts with the result obtained in amphibian embryos injected with anti-FN antibodies (Boucaut et al. 1984). In such treated amphibian embryos, mesoderm fails to migrate into the blastocoel cavity. The ectoderm continues to expand but does not invagi- nate, and no mesoderm forms. However, when the equivalent antibody-blocking experiment is performed in avian embryos, mesoderm does enter from the primitive streak into the underlying space, although it does not migrate laterally (Harrisson et al. 1992). It appears that the result in the FN-null mouse embryo parallels that in the chicken embryo treated with anti-FN antibodies; "mesoderm" forms by delamination from the ectoderm at the primitive streak. Thus, FN is not re- quired for delamination, a process presumably involving reduction in cell-cell adhesion. We are currently analyzing whether or not mesodermal migration is defective in the mutant embryos. Furthermore, in both these cases (mouse and chicken), it is unclear whether the cells that delaminate are true mesoderm. In each case, they are defined by position only, not by expression of particular mesodermal markers. Work is in progress to investigate this issue.

FN-null mouse embryos thus initiate the process of gastrulation in the absence of FN. However, from that point on they become progressively abnormal and are resorbed by day 10-11. Some folding of epithelial sheets occurs, but the embryos are disorganized and many structures fail to form normally, clearly indicating a requirement for FN in morphogenetic processes. Fur-




GENETICS OF CELL-MATRIX A D H E S I O N 255

Figure 6. Comparison of wild-type and homozygous FN-nuU primitive streak stage embryos. (a) Wild-type embryos, stained for FN. FN is concentrated between the three germ layers and within the mesodermal layer. FN is also present in the chorion and between the ectodermal and mesodermal cell layers of the amnion. (b) Homozygous FN-null littermate, stained for FN. No FN is detectable in the embryo, yet three germ layers are present. As for the wild-type littermate, FN is detected in the maternal decidual tissue.

ther analysis of these embryos will provide further insights into the functions of FN during early development. The parallel analysis of the a 5 mutants will similarly define the functions of this integrin subunit.

Although the FN mutants clearly show the importance of FN for development, they provide no information about differential roles for the different isoforms. To address this issue, we have initiated two sets of experiments. In the first, we are attempting to generate more subtle mutations in the FN gene, which are in- tended to alter the pattern of splicing. In the second approach, we are attempting to generate lines of mice expressing FN transgenes of defined splicing pattern; the aim is to breed these transgenic lines with the FN-null line to ascertain which isoforms of FN can rescue the FN-null mutant phenotype and to what degree.

Both approaches have progressed to some extent, although neither is yet complete. Figure 5b illustrates one approach to the generation of FN mutations al- tered in splicing of the B segment, which lies in the center of the 70-kb FN gene. Two mutant constructs

were made, one lacking the exon encoding the B segment and one in which a three-exon/two-intron segment of the gene was replaced by the equivalent segment of cDNA. Each construct also contained a neo R cassette in an intron and a TK cassette at one end (Fig. 5b). ES cell clones with each of these mutations were obtained as described above, and chimeras were derived. Germ-line transmission was obtained in each case, and both heterozygous mutant lines were viable. However, in each case, homozygosity resulted in embryonic lethality, Abnormalit ies developed at about the same time as in the case of FN-null homozygotes, and immunohistochemical analyses indicated that little if any assembled FN was present in these embryos. These data raise the question of whether these mutations in the EIIIB region of the gene lead to more severe defects in FN expression than does simply altering the pattern of splicing, This is now under study. If these replacement constructs prove to be inadequate, an alternative approach will be to use the so-called "hit and run" method (Hasty et al. 1991), which involves insertion of a mutation and subsequent deletion of the adja-




256 HYNES ET AL.

l l k b 7 . 5 k b mouse FN rat FN p r o m o t e r cDNA

1--

T B+ A+ V+ [ B+ A- V+

first intron B- A+ V+ poly A site B - A - V +

Figure 7. FN hybrid transgenes. Defined splicing patterns are listed below the rat cDNA portion of the transgene (stippled box). The first exon of the mouse FN gene is depicted by a white box. Fusion between mouse and rat sequences is in the first intron, depicted by a slanted line. Mice harboring these transgenes will first be analyzed for dominant effects of the transgene. Second, breeding with FN-null mutation mice will determine which FN splice variants are able to rescue the FN-null embryonic lethal phenotype.

cent wild-type segment. This technique has the poten- tial advantage that no neo R cassette is left in the mutated allele.

The transgenic approach we have used is to introduce a hybrid transgene comprising 11 kb of mouse FN promoter and 5'-flanking region fused within the first intron to a rat FN cDNA of defined splicing pattern (Fig. 7). Seven founder lines have been generated and are being tested for expression of rat FN. If these lines produce sufficient FN, they will be intercrossed with the F N - n u l l mice to test for rescue of the embryonic lethality. Since the FN gene is so large ( ~ 70 kb, - 50 introns), we have omitted most of the introns. This may compromise expression. If the current FN transgenic lines prove inadequate, it will probably be necessary to include more of the introns and, perhaps, more flanking DNA.

CONCLUSIONS

Cell-matrix adhesion is complex, involving as it does multiple receptors and multiple ligands, both of which can sometimes undergo alternative splicing to generate isoforms that likely perform different functions. Ge- netic analyses represent one of the best available approaches to elucidation of the contributions of specific receptors or ligands to given functions in vivo. Such studies have been initiated in insects, where the genetics are relatively tractable. However, many of the roles of cell-matrix adhesion arise in important physiological and pathological processes such as hemostasis, thrombosis, inflammation, wound healing, and cancer, which need to be studied in mammalian systems. The development of methods for introducing genes and

Table 2. Murine Mutations in Fibronectin and in Fibronectin Receptor Integrin Subunits

Mutation Heterozygotes Homozygotes

FN - null viable, fertile embryonic lethal FN - B viable, fertile embryonic lethal FN - B § viable, fertile embryonic lethal a 5 - null viable, fertile embryonic lethal cq - null viable ?

defined mutations in mice opens the way to analyses of cell-matrix adhesion by molecular genetics in a mammalian system. We have initiated such an approach to study the functions of FN and integrins that bind to it. The early results indicate that these molecules are in- deed crucial for embryonic development (Table 2). The defects occurring in the mutants are beginning to provide further information about the roles of these molecules, and further analyses using transgenes and more subtle mutations should extend these studies further.

ACKNOWLEDGMENTS

We are grateful to Colleen Leslie for manuscript preparation. This work was supported by grants from the National Cancer Institute (RO1-CA-17007) and the National Heart Lung and Blood Institute (PO1-HL- 41484) and by the Howard Hughes Medical Institute.

REFERENCES

Albelda, S.M. and C.A. Buck. 1990. Integrins and other cell adhesion molecules. FASEB J. 4: 2868.

Bennett, V.D., K.M. Pallante, and S.L. Adams. 1991. The splicing pattern of fibronectin mRNA changes during chondrogenesis resulting in an unusual form of the mRNA in cartilage. J. Biol. Chem. 266: 5918.

Bevilacqua, M., E. Butcher, B. Furie, B. Furie, M. Gallatin, M. Gimbrone, J. Harlan, K. Kishimoto, L. Lasky, R. McEver, J. Paulson, S. Rosen, B. Seed, M. Siegelman, T. Springer, L. Stoolman, T. Tedder, A. Varki, D. Wagner, I. Weissman, and G. Zimmerman. 1991. Selectins: A family of adhesion receptors. Cell 67: 233.

Boucaut, J.C., T. Darribere, H. Boulekbache, and J.P. Thiery. 1984. Prevention of gastrulation but not neurula- tion by antibodies to fibronectin in amphibian embryos. Nature 307: 364.

Brower, D.L. and S.M. Jaffe. 1989. Requirement for integrins during Drosophila wing development. Nature 342: 285.

Brown, N.H., D.L. King, M. Wilcox, and F.C. Kafatos. 1989. Developmentally regulated alternative splicing of Drosophila integrin PS2 a transcripts. Cell 59: 185.

Burton-Wurster, N., G. Lust, and R. Wert. 1989. Expression of the ED B fibronectin isoform in adult human articular cartilage. Biochem. Biophys. Res. Commun. 165: 782.

Capecchi, M.R. 1989. Altering the genome by homologous recombination. Science 244: 1288.




G E N E T I C S O F C E L L - M A T R I X A D H E S I O N 257

Carnemolla, B., E. Baiza, A. Siri, L. Zardi, M.R. Nicotra, A. Bigotti, and P.G. Natali. 1989. A tumor-associated fibronectin isoform generated by alternative splicing of messenger RNA precursors. J. Cell Biol. 108: 1139.

DeSimone, D.W., P.A. Norton, and R.O. Hynes. 1992. Iden- tification and characterization of alternatively spliced fibronectin mRNAs expressed in early Xenopus embryos. Dev. Biol. 149: 357.

Doetschman, T.C., H. Eistetter, M. Katz, W. Schmidt, and R. Kemler. 1985. The in vitro development of blastocyst- derived embryonic stem cell lines: Formation of visceral yolk sac, blood islands and myocardium. J. Embryol. Exp. Morphol. 87: 27.

ffrench-Constant, C. and R.O. Hynes. 1988. Patterns of fibronectin gene expression and splicing during cell migration in chicken embryos. Development 104" 369.

-. 1989. Alternatively splicing of fibronectin is temporal- ly and spatially regulated in the chicken embryo. Develop- ment 106: 375.

ffrench-Constant, C., L. Van De Water, H.F. Dvorak, and R.O. Hynes. 1989. Reappearance of an embryonic pattern of fibronectin splicing during wound healing in the adult rat. J. Cell Biol. 109: 903.

Glukhova. M.A., M.G. Frid, B.V. Shekhonin, Y.V. Balabanov, and V.E. Koteliansky. 1990. Expression of fibronectin variants in vascular and visceral smooth muscle cells in development. Dev. Biol. 14t: t93.

Glukhova, M.A., M.G. Frid, B.V. Shekhonin, T.D. Vas- ilevskaya, J. Grunwald, M. Saginati, and V.E. Kotelian- sky. 1989. Expression of extra domain A fibronectin sequence in vascular smooth muscle cells is phenotype dependent. J. Cell Biol. 109: 357.

Grinblat, Y., S. Zusman, G. Yee, F. Kafatos, and R.O. Hynes. 1992. The role of the cytoplasmic domain of the Drosophila PS/3 integrin during development. Develop- ment (in press).

Grumet, M. 1991. Cell adhesion molecules and their sub- groups in the nervous system�9 Curr. Opin. Neurobiol. 1: 370.

Guan, J.-L. and R.O. Hynes. 1990. Lymphoid cells recognize an alternatively spliced segment of fibronectin via the integrin receptor aJ3~. Cell 60: 51.

Guan, J.-L., J.E. Trevithick, and R.O. Hynes. 1990�9 Ret- roviral expression of alternatively spliced forms of rat fibronectin. J. Cell Biol. 110: 833.

Gulcher, J.R., D.E. Nies, M.J. Alexakos, N.A. Ravikant, M.E. Sturgill, L.S. Marton, and K. Stefansson. 1991. Structure of the human hexabrachion (tenascin) gene. Proc. Natl. Acad. Sci. 88: 9438.

I-Ianahan, D. 1989. Transgenic mice as probes into complex systems. Science 246: 1265.

Harrisson, F., J. van Hoof, and J.-M. Foidart. 1992. Mi- croinjection of anti-fibronectin antibodies in the chicken blastoderm: Inhibition of mesoblast cell migration but not of cell ingression at the primitive streak�9 Development (in press).

Hasty, P., R. Ramirez-Solis, R. Krumlauf, and A. Bradley. 1991. Introduction of a subtle mutation into the Hox-2.6 locus in embryonic stem cells. Nature 350: 243.

Hay, E.D., ed. 1991. Cell biology ofextracellular matrix, 2nd edition. Plenum Press, New York.

Hemler, M.E. 1990. VLA proteins in the integrin family: Structures, functions and their role on leukocytes. Annu. Rev. lmmunol. 8: 365.

Hogan, B., F. Costantini, and E. Lacy. 1986. Manipulating the mouse embryo�9 A laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.

Hynes, R.O. 1987. Integrins: A family of cell surface receptors. Cell 4: 549.

�9 1990. Fibronectins. Springer-Verlag, New York. �9 1992. Integrins: Versatility, modulation, and signaling

in cell adhesion. Cell 69: 11. Hynes, R.O. and A.D. Lander. 1992. Contact and adhesive

specificities in the associations, migrations, and targeting of cells and axons. Cell 68: 303.

Jessell, T.M. 1988. Adhesion molecules and the hierarchy of neural development�9 Neuron 1: 3.

Knowlton, A.A . , C.M. Connelly, G.M. Romo, W. Mamuya, C.S. Apstein, and P. Brecher. 1992. Rapid expression of fibronectin in the rabbit heart after myocardial infarction with and without reperfusion. J. Clin. Invest�9 89: 1060.

Lasky, L.A. and S. Rosen. 1992. The selectins: Carbohydrate- binding adhesion molecules of the immune system�9 In Inflammation: Basic principles and clinical correlates (ed. J. Gallin et al.). Raven Press, New York. (In press.)

Leptin, M., T. Bogaert, R. Lehmann, and M. Wilcox. 1989. The function of PS integrins during Drosophila em- bryogenesis. Cell 56: 401.

Mamuya, WS. and P. Brecher. 1992. Fibronectin expression in the normal and hypertrophic rat heart. J. Clin. Invest. 89: 392.

Mansour, S.L., K.R. Thomas, and M.R. Capecchi. 1988. Disruption of the proto-oncogene int-2 in mouse embryo- derived stem ceils: A general strategy for targeting mutations to non-selectable genes. Nature 336: 348.

Mould, A.P., A. Komoriya, K.M. Yamada, and M.J. Humph- ries. 1991. The CS5 peptide is a second site in the IIICS region of fibronectin recognized by the integrin a4/31. Inhi- bition of oq/31 function by RGD peptide fiomologues. J. Biol. Chem. 266: 3579.

Mould~ A.P_ L.A. Wheldon, A. Komoriya, E.A. Wayner, K.M. Yamada, and M.J. Humphries. 1990. Affinity chro- matographic isolation of the melanoma adhesion receptor for the IIICS region of fibronectin and its identification as the integrin C r a f t 1 . J. Biol. Chem. 265: 4020.

Norton, P.A. and R.O. Hynes. 1987. Alternative splicing of chicken fibronectin in embryos and in normal and trans- formed cells. Mol. Cell Biol. 7: 4297.

Oyama, F., S. Hiroshashi, Y. Shimosato, K. Titani, and K. Sekiguchi. 1989. Deregulation of alternative splicing of fibronectin pre-mRNA in malignant human liver tumors�9 J. Biol. Chem�9 264: 10331.

Pagani, F., L. Zagato, C. Vergani, G. Casari, A. Sidoli, and F.E. Baralle. 1991. Tissue-specific splicing pattern of fibronectin messenger RNA precursor during development and aging in rat. J. Cell Biol. 113: 1223.

Patel, R.S., E. Odermatt, J.E. Schwarzbauer, and R.O. Hynes. 1987. Organization of the rat fibronectin gene pro- vides evidence for "exon shuffling" during evolution. EMBO J. 6: 2565.

Paul, J.I., J.E. Schwarzbauer, J.W. Tamkun, and R.O. Hynes. 1986. Cell-type-specific fibronectin subunits generated by alternative splicing. J. Biol. Chem. 261: 12258.

Peters, J.H., M.H. Ginsberg, C.M. Case, and C.G. Coch- rane. 1988. Release of soluble fibronectin containing an extra type Ill domain (ED1) during acute pulmonary injury mediated by oxidants or leukocytes in vivo. Am. Rev. Respir. Dis. 138: 167.

Schwarzbauer, J.E., C.S. Spencer, and C.L. Wilson. 1989. Selective secretion of alternatively spliced fibronectin variants. J. Cell Biol. 109: 3445.

Schwarzbauer, J.E., R.S. Patel, D. Fonda, and R.O. Hynes. 1987. Multiple sites of alternative splicing of the rat fibronectin gene transcript. EMBO J. 6: 2673.

Shattil, S.J. and J.S. Brugge. 1991. Protein tyrosine phosphor- ylation and the adhesive functions of platelets. Curr. Opin. Cell Biol. 3: 869.

Shekhonin, B.V., S.B. Guriev, S.B. lrgashev, and V.E. Koteliansky. 1990. Immunofluorescent identification of fibronectin and fibrinogen/fibrin in experimental myocardial infarction. J. Mol. Cell. Cardiol. 22: 533.

Siri, A., B. Carnemolla, M. Saginati, A. Leprini, G. Casari, F. Baralle, and L. Zardi. 1991. Human tenascin: Primary structure, pre-mRNA splicing patterns and localization of the epitopes recognized by two monoclonal antibodies. Nucleic Acids Res. 19: 525.




258 H Y N E S E T AL.

Spring, J., K. Beck, and R. Chiquet-Ehrismann. 1989. Two contrary functions of tenascin: Dissection of the active sites by recombinant tenascin fragments. Cell 59: 325.

Takeichi, M. 1988. The cadherins: Cell-cell adhesion molecules controlling animal morphogenesis. Development 102: 639.

- - . 1990. Cadherins: A molecular family important in selective cell-cell adhesion. Annu. Rev. Biochem. 59: 237.

�9 1991. Cadherin cell adhesion receptors as a morphogenetic regulator. Science 251: 1451.

Thomas, K.R. and M.R. Capecchi. 1987. Site-directed muta- genesis by gene targeting in mouse embryo-derived stem cells. Cell 51: 503.

Vartio, T., L. Laitinen, O. Narvanen, M. Cutolo, L.E. Thor- nell, L. Zardi, and I. Virtanen. 1987. Differential expression of the ED sequence-containing form of cellular fibronectin in embryonic and adult human tissues. J. Cell Sci. 88: 419.

Wartiovaara, J., I. Leivo, and A. Vaheri. 1979. Expression of

the cell surface-associated glycoprotein, fibronectin, in the early mouse embryo�9 Dev. Biol. 69: 247.

Wayner, E.A., A. Garcia-Pardo, M.J. Humphries, J.A. McDonald, and W.G. Carter. 1989. Identification and characterization of the T lymphocyte adhesion receptor for an alternative cell attachment domain (CS-1) in plasma fibronectin. J. Cell Biol. 109: 1321.

Wilcox, M. 1990. Genetic analysis of the Drosophila PS integrins. Cell Differ. Dev. 32: 391.

Wilcox, M., A. DiAntonia, and M. Leptin. 1989. The function of PS integrins in Drosophila wing morphogenesis. De- velopment 107: 891.

Zetter, B.R. and G.R. Martin�9 1978. Expression of a high molecular weight cell surface glycoprotein (LETS protein) by preimplantation mouse embryos and teratocarcinoma stem cells�9 Proc. Natl. Acad. Sci. 75: 2324.

Zusman, S., Y. Grinblat, G. Yee, F.C. Kafatos, and R.O. Hynes. 1992. Analyses of PS integrin functions during Drosophila development. Development (in press).




toward a genetic analysis of cell-matrix adhesion - biologybiology.hunter.cuny.edu/cellbio/feinstein...

Documents