Steady-state levels of mRNAs coding for the type IV collagen and laminin polypeptide chains of basement membranes exhibit marked tissue-specific stoichiometric variations in the rat.

Rat retina, lens, and kidney from 8-week-old animals were assayed for the steady-state levels of mRNAs for four basement membrane components: The alpha 1 chain of type IV collagen, the alpha 2 chain of type IV collagen, the B1 chain of laminin, and the B2 chain of laminin. Each tissue exhibited markedly different ratios of the four mRNAs. The mRNA ratio for the alpha 1 chain of type IV collagen to the B1 chain of laminin varied from a value of 0.7 in retina to a value of 17 in lens. Also, the mRNA ratio for the alpha 1 chain to the alpha 2 chain of type IV collagen varied from 1.6 in retina to 17 in lens, and the mRNA ratio for the B1 chain to the B2 chain of laminin varied from 0.6 in lens to 2.9 in kidney. The mRNA coding for the alpha 1 chain of type IV collagen decreased in all three tissues as the animals increased in age from 8 to 16 weeks, with the rate of decline being greater in retina than in lens of kidney. The levels of mRNA coding for the B1 and the B2 chains of laminin decreased in the kidney between 8 and 16 weeks but at different rates. Comparison of mRNAs from kidney of rats over this time period showed that the ratio of alpha 1 to B1 remained relatively constant with age, whereas the ratio of B1 to B2 increased. One possible explanation for the results is that each tissue has elaborate, tissue-specific controls for translation that provide synthesis of basement membrane components in the same proportion, in spite of the varying steady-state levels of the mRNAs. A more likely explanation is that different tissues synthesize type IV collagen and laminin at different rates, and that even the subunit compositions of the type IV collagen and laminin molecules vary from tissue to tissue and in an age-dependent manner.

Rat retina, lens, and kidney from 8-week-old animals were assayed for the steady-state levels of mRNAs for four basement membrane components: The a1 chain of type IV collagen, the a2 chain of type IV collagen, the B1 chain of laminin, and the B2 chain of laminin. Each tissue exhibited markedly different ratios of the four mRNAs. The mRNA ratio for the a1 chain of type IV collagen to the B1 chain of laminin varied from a value of 0.7 in retina to a value of 17 in lens. Also, the mRNA ratio for the a1 chain to the a2 chain of type IV collagen varied from 1.6 in retina to 17 in lens, and the mRNA ratio for the B1 chain to the B2 chain of laminin varied from 0.6 in lens to 2.9 in kidney. The mRNA coding for the a1 chain of type IV collagen decreased in all three tissues as the animals increased in age from 8 to 16 weeks, with the rate of decline being greater in retina than in lens or kidney. The levels of mRNA coding for the B1 and the B2 chains of laminin decreased in the kidney between 8 and 16 weeks but at different rates. Comparison of mRNAs from kidney of rats over this time period showed that the ratio of a1 to B1 remained relatively constant with age, whereas the ratio of B1 to B2 increased. One possible explanation for the results is that each tissue has elaborate, tissue-specific controls for translation that provide synthesis of basement membrane components in the same proportion, in spite of the varying steady-state levels of the mRNAs. A more likely explanation is that different tissues synthesize type IV collagen and laminin at different rates, and that even the subunit compositions of the type IV collagen and laminin molecules vary from tissue to tissue and in an age-dependent manner.
The investigation of BM' metabolism in vivo has to date been based largely on studies examining rates of radiolabel inclusion and turnover (1-3), assays of enzymes involved in collagen post-translational modification (4-6), or measurement of serum levels of BM-derived antigens (7). These * This work was supported in part by National Institutes of Health Grants AM 16516 and GM 34090, The Lions Clubs International Research Program of the American Diabetes Association, and the National Society to Prevent Blindness (New York Affiliate). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
11 To whom correspondence should be addressed. 'The abbreviations used are: BM, basement membranes; SDS, sodium dodecyl sulfate; kb, kilobases. somewhat limited approaches were used because of the difficulty of isolation and inherent insolubility of BMs, and because of the very low levels of BM synthesis in most tissues.
The identification and characterization of cDNAs coding for polypeptide components of BM, such as the al(1V) and a2(IV) chains of type IV collagen (8)(9)(10)(11)(12) and the laminin B1 and B2 chains (13), has provided a highly sensitive and specific means by which BM synthesis can be examined concurrently in different tissues, namely, by assaying the steady-state tissue concentrations of the mRNAs coding for these matrix proteins.
During the course of investigations to determine the effects of diabetes on BM synthesis in the rat (14), it became apparent that the retina, lens, and kidney exhibited marked but specific stoichiometric variations in the relative steady-state levels of these BM-coding mRNAs. This investigation addresses the quantitation of these tissue-specific variations and the changes in these mRNA levels that occur with age. MATERIALS AND METHODS RNA Isolation-Total RNA was isolated from the retina, lens, and kidney of male CDF rats (Charles River Laboratories, Wilmington, MA) by the guanidinium thiocyanate/cesium chloride method (15). RNA was shown to be intact by the presence of discrete 28 S and 18 S ribosomal bands after electrophoresis in formaldehyde (2.2 M) containing 1% agarose gels stained with ethidium bromide (15).
cDNA Probes and Labeling-The following mouse cDNAs were utilized in this investigation: pPE123 containing E 1.8-kb insert coding for the 204 amino acids at the C terminus of the NCl domain of the al(IV) collagen chain and 1.2 kb of 3'-untranslated region ( l l ) , pPE18 containing a 1.1-kb insert coding for approximately 370 amino acids located within the central triple helix-forming region of the a2(IV) collagen chain (12), pPE386 containing a 1.1-kb insert coding for approximately 300 amino acids at the C terminus of the laminin B1 chain and 116 bases of 3"untranslated region (13), and pPE9 containing a 675-base insert coding for the C-terminal end of the laminin B2 chain and 20 bases of 3"untranslated region (13).
The inserts were removed from the described plasmids by restriction endonuclease digestion and routinely labeled with [32P]dCTP by nick translation to specific activities of 4-9 X l O a cpm/pg of DNA. Northern and Slot-blotting Procedures-Northern analysis was performed by electrophoresis of heat-denatured RNA in a formaldehyde (2.2 M)-containing 1% agarose gel followed by capillary blotting onto nitrocellulose (15). For slot-blotting, known quantities of total RNA, assessed by absorbance at 260 nm, were denatured by heating in RNase-free HZ0 at 68°C for 15 min. The solutions were then rapidly chilled on ice, serially diluted, and applied to nitrocellulose filters in a denaturing buffer (6 X SSC, 2.5 M formaldehyde (1 X SSC = 0.15 M NaCl, 0.015 sodium citrate, pH 6.8)) utilizing a slot-blotting apparatus (Minifold 11, Schleicher & Schuell).
Hybridization Conditions-Northern and slot-blotted filters were air-dried, baked under vacuum at 80 "C for 2 h, and prehybridized in 50% formamide, 4 X SSC, 1 mg/ml sheared salmon sperm DNA, 1 X Denhardt's solution (14), and 0.1% SDS for 6 h at 37 'C. Hybridiza-tions with heat-denatured '*P-labeled cDNAs were conducted at 37 "C for 24 h in the same solution as described for prehybridizations.
Filters were washed to a final stringency of 0.2 X SSC and 0.1% SDS at 52 "C.
Autoradiograms were generated by exposure of the filters to x-ray film (Kodak XAR) in the presence of intensifying screens at -80°C. The bound label was assayed by absorbance per microgram of RNA from densitometiic analysis of the autoradiographs.
Statistics-Statistical analyses were performed by the two-tailed Student's t test.

RESULTS
The Northern analysis of rat parietal endoderm total RNA ( Fig. 1) demonstrates that, under the described hybridization and washing conditions, the mouse-derived cDNA inserts hybridize solely and specifically with their rat mRNA counterparts.
Autoradiographs obtained from four identical slot-blots of total RNA obtained from the tissues of 8-week-old rats were hybridized with cDNAs coding for al(1V) and a2(IV) chains of type IV collagen, and the B1 and B2 chains of laminin (Fig.   2). Changes in absorbance per microgram of RNA obtained from densitometric analysis of the autoradiographs together with stoichiometric comparisons are presented in Table I. It should be noted that, since the cDNAs utilized varied in length (0.6-1.8 kb) and specific activity (4-9 x 10' cpmlpg of DNA), the ratios of mRNA levels presented are on an arbitrary rather than a molar scale.
The lens and kidney exhibited steady-state levels of mRNA coding for al(1V) chain that were on the order of 4 times higher per microgram of total RNA than that found in the retina ( Fig. 2 and Table I). However, the steady-state levels of a2(IV) mRNA in lens and kidney were about half those of retina. Therefore, there was a markedly lower ratio of al:a2(IV) chain mRNA in retina (Table I). All results presented were checked by reprobing the four slot-blots with the cDNAs in a different order after removal of the original label by treatment with 50% formamide and 0.1 x SSC a t 65 "C for 4 h.
The kidney contained twice as much mRNA coding for B1 laminin per microgram of total RNA as did the retina, and more than 10 times as much as the lens ( Fig. 2 and Table I).
Laminin B2 mRNA analysis indicated that the kidney again Northern analysis of rat parietal endoderm total RNA hybridized with "P-labeled mouse-derived BM-coding cDNAs. Total RNA (5 pg) from rat parietal endoderm was electrophoresed in a formaldehyde-containing agarose gel, blotted, hybridized, washed, and exposed to x-ray film under the same conditions subsequently used for slot-blotting (see "Materials and Methods"). The same filter was successively probed in the following order: 1) PE386 (laminin Bl), 2) PE18 (a2(IV) collagen), 3) PE123 (aI(1V) collagen), and 4) PE9 (laminin B2). In each case, the previous 32Plabeled cDNA was removed by incubating the filter in 0.1 X SSC containing 50% formamide at 65°C for 4 h. The positions of the 28 S and 18 S ribosomal RNA bands are indicated as size markers. Autoradiographs of slot-blotted total RNAs hybridized to s'P-labeled cDNAs for type IV collagen and laminin. Total RNA extracted from retina, lens, and kidney was slot-blotted on nitrocellulose filters in the amounts indicated. The four identical blots were hybridized with nick-translated cDNAs for the two chains of type IV collagen (PE123:al chain, PE18a2 chain) and two of the laminin chains (PE386B1, PE9B2). After washing, the filters were exposed to x-ray film for 24 h as described under "Materials and Methods."

TABLE I
Relative steady-state levels of mRNAs for a1 and a2 chains of type IV collagen and the Bl and 8 2 chins of laminin at 8 weeks of age To illustrate the variations from tissue to tissue, the results were calculated from a single autoradiograph prepared with equal amounts of RNA from the three tissues (see Fig. 2). Values for mRNA are calculated from densitometric analysis of the radiograph and are expressed in arbitrary absorbance units per pg of RNA. As indicated in Tables I1 and 111, the S.E. for the assays is about *30% of the mean. As indicated in the text, approximately the same values for the ratio of mRNAs were obtained when the same filters were reprobed in a different order with a new set of nick-translated cDNA mobes.  Table I).
The relative levels of type IV collagen to laminin mRNAs in the lens were far greater than that exhibited by either the retina or the kidney (Table I). On the basis of the 'data presented in Table I, the retina appears to synthesize BM with the lowest levels of type IV collagen relative to laminin B chains.
Having established that each of the three tissues examined in 8-week-old rats exhibited its own specific pattern of steadystate levels of BM-coding mRNAs, we proceeded to investigate how these levels change as the rat increases in age over the period of 8-16 weeks. Table I1 presents   The total RNA samples extracted from a particular tissue at each time point were slotted concurrently on one set of nitrocellulose filters. This set of filters was then prehybridized, hybridized, and washed under identical conditions. The set was then exposed to x-ray film for a fixed period (24-48 h, depending on tissue) and the mRNA levels quantified as described under "Materials and Methods." p values are a: whereas the retinal level was reduced to about one-tenth over the same time period (Table 11). Levels of the mRNA for the a2(IV) chain were too low for accurate assay in animals older than 8 weeks.
In the same animals, measurement of kidney levels of B1 and B2 laminin mRNA revealed that these mRNAs also decrease with age (Table 111). However, the decreases observed did not appear to be closely coordinated, in that the B2 laminin mRNA level was reduced to 26% over this 8-week period, whereas the B1 mRNA level fell to only 72% (Table  111).
Assay of levels of mRNA for @-actin in the same three tissues indicated that the levels did not fall significantly in retina or lens between 8 and 16 weeks. The values at 16 weeks were 83 f 12% of the 8-week value for retina and 80 f 8% of the 8-week value for lens. The level of P-actin decreased progressively in kidney so that at 16 weeks the value was 58 f 10% of the 8-week value.
A comparison of the ratio of mRNAs coding for the al(1V) collagen to the B1 laminin chain, and of the B1 to B2 laminin chains in the kidney over the 8-16-week period is presented in Fig. 3. Whereas the a1:Bl ratio remained relatively constant over this period and showed no consistent trend, the Bl:B2 laminin ratio exhibited a marked increase (Fig. 3). At 16 weeks, the kidney Bl:B2 mRNA ratio was increased 2.8fold in comparison with the ratio obtained at 8 weeks of age.

DISCUSSION
Heterogeneity among BMs from different tissues was previously suggested on the basis of amino acid composition, polypeptide profiles on SDS-polyacrylamide gels, and immunochemical analysis (15)(16)(17). However, it has been difficult to confirm this suggestion because the individual components cannot be extracted in high yield from most BMs and because BMs cannot be isolated in pure form from most tissues. Also, it is generally accepted that the type IV collagen molecule contains two a1 chains and one a 2 chain, whereas the laminin molecule contains one B1 chain, one B2 chain, and one A chain. However, the chain composition of the type IV molecule has not been fully resolved by chemical analysis, and Haralson et al. (19) recently demonstrated synthesis of (~1 ( 1 V )~ collagen molecules in cultured parietal yolk sac cells. Similarly, the structure of the large laminin molecule has not been fully defined, and noncoordinate synthesis of the laminin B and A chains was demonstrated previously in mouse eggs and early embryos (20) as well as during the differentiation of F9 embryonal carcinoma cells (21).
The results obtained from 8-week-old rats establish that the ratio of the steady-state levels of mRNAs for type IV collagen to laminin vary over a 20-fold range among retina, lens, and kidney. In addition, they demonstrate that the ratio of the steady-state levels of mRNAs for the two polypeptide chains of type IV collagen and for two of the chains of laminin vary over a 5-10-fold range in the same tissues. These differences were tissue-specific and also dependent on age. The a1 type IV collagen mRNA level in retina decreased with age at a greater rate than that noted in lens or kidney. Furthermore, kidney B1 and B2 laminin mRNA levels also decreased with age but at markedly different rates. This resulted in a steady increase of the ratio of B1 to B2 laminin mRNA levels in this tissue with age, whereas the al(1V) collagen to B1 laminin mRNA ratio exhibited no consistent trend.
Steady-state levels of mRNA do not always reflect rates of protein synthesis. Therefore, it is possible that the large variations seen here in the ratios of mRNAs are compensated for by differences in the rates of translation so as to generate the components in the same stoichiometry in all tissues. This, however, would require an elaborate series of controls for translation of specific mRNAs. Therefore, the simplest explanation for the data presented herein is that the tissues synthesize the matrix components at different rates and that the BMs contain different amounts of type IV collagen and laminin. Also, the data are consistent with the conclusion that the chain composition of both the type IV collagen molecule and laminin molecule varies from tissue to tissue. Furthermore, the kidney data suggest that the chain composition of laminin may vary within a particular tissue as a function of age. Thus, tissues appear to regulate their BM-coding mRNA levels in such a way as to allow the synthesis and composition 10. Pihlajaniemi