J. Embryo!, exp. Morph. Vol. 56, pp. 225-238, 1980 2 2 5Printed in Great Britain © Company of Biologists Limited 1980

Analysis of glycosaminoglycans duringchondrogenesis of normal and brachypod

mouse limb mesenchyme

By JANET SHAMBAUGH1 AND WJLLIAM A.ELMER2

From the Biology Department, Emory University, Atlanta, Georgia

SUMMARYStudies have been carried out to characterize radioactive incorporation rates and steady-state levels of hyaluronic acid (HA) and chondroitin sulfate (CS) as well as hyaluronidaseactivity in the hind limbs of normal and brachypod mouse embryos between 11-14 days ofgestation. The results of the analysis show that changes in the synthetic and degradative ratesof HA and CS occur at about the 12-5-day stage in normal hind limbs. These changes includean increased rate of CS synthesis, a decreased rate of HA synthesis, and a correspondinglysharp, transitory rise in hyaluronidase activity. Similar changes also occur in brachypod hindlimbs but appear to be delayed in onset by approximately one-half day. In addition, themutant hind limb exhibits a slower loss of HA at a time when turnover would be expectedto be on the increase. These changes are concomitant with cell surface alterations and ab-normal mesenchymal condensation formation which have been previously shown to occurin this mutant.

INTRODUCTION

Changing physiological levels of the cartilage glycosaminoglycans (GAGs),hyaluronic acid (HA) and chondroitin sulfate (CS), are thought to be importantto chondrogenesis, and therefore, to normal skeleton formation. Thus, a corre-lation has been shown between HA and morphogenesis in several developingsystems (Polansky, Toole & Gross, 1973; Toole, 1973 a; Pratt, Larson & John-ston, 1975; Solursh, 1976; Greenberg & Pratt, 1977; Solursh & Morriss, 1977;Derby, 1978, Orkin & Toole, 1978), and removal of HA has been correlated withsubsequent cytodifferentiation of chondrocytes (Toole, 1972; Corsin, 1975;Smith, Toole &~ Gross, 1975). While most studies of chondrogenesis havefocused on CS or CS proteoglycan in the chick limb system (Goetinck, Penny-packer & Royal, 1974; Hascall, Oegema, Brown & Caplan, 1976; DeLuca et al.1977; Holtzer, Okayama, Biehl & Holtzer, 1978) and in the mouse limb system(Royal & Goetinck, 1977), changing levels of both GAGs have been describedusing only radioactive labeling in chick limb development (Toole, 1972) andnewt limb regeneration (Toole & Gross, 1971; Smith et al. 1975). A few studieshave measured actual levels of HA and CS - one in chick axial skeleton (Kvist& Finnegan, 1970) and another in neurulating chick (Derby, 1978).

1 Author's address: Biology Department, Johns Hopkins University, Baltimore, Maryland21218 U.S.A.

2 Author's address: Biology Department, Emory University, Atlanta, Georgia 30322, U.S.A.

226 J. SHAMBAUGH AND W. A. ELMER

The present study characterizes radioactive incorporation rates and steady-state levels of HA and CS, as well as hyaluronidase (HAdase) activity, duringthe early stages of hind-limb development in normal ( + / + ) mice and micewhich carry the brachypod (bp^/bp3) mutation. This mutation, which affectsthe differentiation and development of the appendicular skeleton, appears toact by causing a delay and reduction in the formation of the mesenchymalcondensations which subsequently develop into the cartilaginous anlagen of theskeletal elements (Griinebarg & Lee, 1973; Elmer, 1976). The defect in the con-densation process has been correlated with alterations in the surface propertiesof the prechondrogenic mesenchyme (Duke & Elmer, 1977, 1978, 1979; Hewitt& Elmer, 1978). It has not been reported, however, whether biochemical differ-ences of extracellular matrix components are being displayed concomitantlyduring this developmental time period. The data presented in this paper indicatethat an alteration in the levels of the glycosaminoglycans is also occurring at thistime period.

MATERIALS AND METHODSSource of tissue

Normal and mutant mice were obtained from Jackson Laboratory, Bar Har-bor, Maine. Homozygous bp3 /bp3 mice were bred onto the same hybrid back-ground as the phenotypically normal mice ^CAFi/J). Embryos of a particulargestational age were generated using timed matings with the appearance ofthe vaginal plug called day zero. Embryos were removed from the uterus inwarm, sterile Tyrode's solution and staged as described previously (Elmer &Selleck, 1975). The hind-limb buds were excised using iridsctomy scissors andeither placed into organ culture (Kochhar & Aydelotte, 1974) or treated furtherfor cell culture (Elmer & Selleck, 1975). For culturing the whole limb bud, themedium consisted of 75 % BGJ chemically defined medium (Difco Laboratories),25 % fetal calf serum (GIBCO), 150 /tg/ml ascorbate (Nutritional BiochemicalsCorp.), and 50 /tg/ml gentamicin sulfate (Garamycin, Schering PharmaceuticalCorp.). For cell culture, limb mesenchyme was obtained by removing the ecto-derm with microneedles after a 5 min incubation at room temperature in 2-25 %trypsin (GIBCO) plus 0-75 % pancreatin. The resulting mesodermal core wasplaced in calcium-magnesium-free Tyrode's solution for 30 min at 25 °C todissociate the cells. The cells were plated in 35 mm plastic tissue culture dishes(Falcon) at a density of 1 x 106 cells/ml in 3 ml Eagle's Basal Medium withHank's salts supplemented with glutamine (GIBCO), 10% horse serum(GIBCO), and 50 ^g/ml gentamicin sulfate. Both cell and organ cultures wereincubated in a humidified atmosphere of 5 % CO2, 95 % air, at 37 °C.

Quantification of glycosaminoglycans

The radioactive precursor D- [6-3H]glucosamine-HCl (sp. act. 20-30 Ci/mmole, Amersham-Searle) used in the synthesis of HA, CS and glycoproteins

Glycosaminoglycan levels in normal and brachypod limbs 227

(GPs) was added at t0 to the medium at a concentration of 5 /tCi/ml for theorgan cultures and 1 /tCi/ml for the cell cultures. Since the incorporation ratewas observed to be linear for at least 12 h for both genotypes, a 6 h labelingperiod was used throughout all of the experiments. Cell and organ cultures werealso labeled concurrently with 5 /*Ci/ml of Na2

35SO4 (carrier free, Amersham-Searle) as an additional indicator of sulfated GAGs. The GAGs were extractedfrom either one cell-culture plate (cell layer plus medium) or from 10 limb budscombined with 20 unlabeled limb buds by shaking in 0-5 M-NaOH and 1-0 Msodium borohydride overnight at 25 °C. After dialysis against running waterovernight, an aliquot was removed for DNA determination (Giles & Myers,1965). For the limb-bud retentate, the GAGs were precipitated with three vol-umes of cold ethanol plus 5 % potassium acetate, and chromatographed on acolumn of DEAE-52 cellulose (Whatman) with a linear gradient of 0-1-2 M-LiCl in 0-05 M tris buffer, pH 7-2 (Pratt, Goggins, Wilk & King, 1973). Fractionswere collected (3 ml) and a 0-5 ml aliquot of each fraction was assayed forradioactivity in 10 ml of ACS (Amersham-Searle) liquid scintillant using aBeckman LS 100 counter to determine the elution profile. The fractions withineach peak (fractions 14-20 for HA and 22-37 for CS) were pooled, dialyzed,concentrated to 1-0 ml using a Molecular Separator Kit (Millipore) and assayedfor uronic acid content (Bitter & Muir, 1962) and radioactivity. The identityof the peaks was confirmed by digesting the concentrated effluent with enzymesof narrow specificity: l-0mg/ml leech head hyaluronidase (Biotrics, Inc.,Arlington, Mass.), 5 units/ml chondroitinase AC or ABC (Sigma) in enrichedtris buffer (Saito, Yamagata & Suzuki, 1968), and l-0mg/ml heparanase(kindly supplied by A. Linker, V. A. Hospital, Salt Lake City, Utah) in 01 Msodium acetate buffer, pH 7-0 (Linker & Hovingh, 1972).

HA levels in limb-bud extracts were also quantified using the Reissig test(Reissig, Strominger & Leloir, 1955) after digestion with Streptomyces HAdase(Calbiochem, Grade B) as described by Hatae & Makita (1975). For this assaythe NaOH extract was first deproteinated by precipitation with 10% trichloro-acetic acid (TCA). The TCA was extracted with three changes of an equalvolume of ethyl ether, and the GAGs were precipitated from the aqueous phasewith 3 vol. of cold ethanol containing 5 % potassium acetate. The GAGswere dissolved in 0-2 ml of 0-05 M sodium acetate buffer, pH 5-0, and twoTurbidity Reducing Units (TRU) of Streptomyces HAdase were added. Afterincubation overnight at 55 °C, the pH was adjusted to 9-2 before measuring freehexosamine reducing ends.

Cell cultures were extracted in a similar manner. As described above, excessprotein was first removed from the dialysis retentate using TCA. Separation ofGAGs was then accomplished by sequential collection on a filter of pore size0-45 /tm (Millipore) according to methods of Saarni & Tammi (1977). Thescintillant used was ACS (Amersham-Searle).

228 J. SHAMBAUGH AND W. A. ELMER

Loss of glycosaminoglycans in culture

Approximation of GAG turnover in limb-bud organ cultures was measuredby pulse labelling for 6 h followed by a 4-day incubation in medium minusthe radioactive precursor and containing a 50-fold excess (10 fiu) of unlabeledglucosamine. Cultures were fed by changing 1 ml of medium each day. Cultureswere terminated at 1-day intervals by extracting four limb buds as describedabove. The extract was dialyzed, and DNA content and radioactivity determined.

HAdase activity in limb-bud homogenates

HAdase activity was measured in a crude homogenate of 50-100 limb budsby the Reissig test (Reissig et al. 1955) using 100 fig of HA as substrate (Sigma,Umbilical Cord) for a 20 h incubation at 37 °C (Polansky, Toole & Gross,1973). Alternatively, [3H]HA isolated from limb-bud cultures by DEAE chroma-tography was used as substrate. A unit of activity is denned as fig HAdigested/10 limb buds/24 h.

RESULTS

Glycosaminoglycans were quantified over developmental time in two waysin order to obtain a more complete description of the changes occurring inthe limb bud. In one assay, a colorimetric analysis of the GAG levels wascarried out to determine the quantity of GAGs at different stages of earlydevelopment. The other assay involved the use of the radioactive precursors,[3H]glucosamine and Na35SO4, to determine the amount of HA, CS and GPs

Table 1. Deoxyribonucleic acid levels in normal and br achy pod limb buds*

Day of gestation

11-5 12-5 13-5 14-5

9-l±2-4f 17-2±3-2 23-6±0-9 31-3±41bpu/bpK 9-2 ±1-3 17-4 ±3-8 23-2 ±0-8 30-0+1-7No. of determinations 9 12 5 4

* DNA was quantified by the diphenylamide reaction in dialyzed sodium hydroxide ex-tracts of limb buds.

t Mean±s.D.

produced during a particular 6 h period. However, since these values are deter-mined by synthesis minus degradation, the loss of radioactivity was alsomeasured to determine the importance of degradation to rate. The data areexpressed on either a per limb bud or per fig of DNA basis since the DNAcontent was not significantly different between the two genotypes in limb budsof the same age (Table 1).

To perform the colorimetric and radioactive assays, HA, CS and GP were

Glycosaminoglycan levels in normal and brachypod limbs 229

first separated by anion exchange chromatography as described under Methods.A sample chromatograph is shown in Fig. 1, indicating how the fractions werepooled to give three groups. Peak one co-elutes with HA standard, whereas,peak two which contains both 3H and 35S co-elutes with CS standard. The exactnature of the components present in the two major peaks was confirmed usingenzymes with narrow specificities. These results are summarized in Table 2.The first major peak is 90% digestible by leech hyaluronidase. This enzymehas absolute specificity for HA. The second, broader peak is about 40%chondroitin sulfate A and C, 30% chondroitin sulfate B, and 20% heparansulfate. The remaining 10% could be keratin sulfate since this GAG is notdegraded by any of the enzymes used. Since the first major peak contains mainlyHA and the second peak contains mainly CS, HA and CS were routinely quanti-fied by summing the fractions (14-20 for HA and 22-37 for CS) within theirrespective peaks. All fractions collected before the HA peak were also pooledto quantitate the GPs (Pratt et al. 1973). The percent composition of the peaksdid not differ significantly between the two genotypes (Table 2), therefore HAand CS may be compared between genotypes by measuring the material collectedwithin the two peaks.

Table 2. Identification of components in the DEAE columneffluent using specific enzyme digestions

Day 11-5 Day 12-5 Day 13 5 Day 14-5Peak 1 *

% digested by 72 (74)f 88 (92) 90 94 (86)leech HAdase

Peak 2% digested by 58 (49) 65 (68) 75 12 (61)CSase ABCCSase AC — 35 (35) 40 —Heparanase — 17 (16) 19 —Leech HAdase 0(0) — — 0(0)

* Peak 1 (fractions 14-20) represents HA, peak 2 (fractions 22-37) is CS.f Values in parentheses are results from digestion of pb^/bp™. All other values are from

+ / + limb mesenchyme extracts.

Chondroitin sulfate measurements

The radioactive incorporation of [3H]glucosamine into CS (Fig. 2 A) displaysa constant rate until 12-5 days of gestation, when the rate increased linearly(r2 = 0-97 for linear regression of line described by last four points, n = 2). Thetiming of this abrupt change corresponds to the onset of visible areas of cellcondensation and cartilage matrix deposition in the normal limb. In the mutant,condensations are also visible, however, they are smaller in size (Griineberg &Lee, 1973) and contain a lesser amount of matrix material (Duke & Elmer,

230 J. SHAMBAUGH AND W. A. ELMER

45

x 30£

15

Glycoproteins 'Peak one Peak two

12 18 24Fraction number

30 36

Fig. 1. A representative DEAE-cellulose chromatogram of glycosaminoglycansfrom 13-day + / + mouse hind limbs. [3H]Glucosamine and Na2

35SO4, each at5 /iCi/ml, were added at t0 to the organ culture medium for a period of 6 h. Ten limbbuds were extracted together with 20 unlabeled buds, and the GAGs precipitated withethanol before being applied to the column. Fractions were tested for radioactivity,and the material within each peak was pooled as indicated by the vertical lines. Theconcentrated peaks were tested for uronic acid content and enzyme susceptibility.Peak one represents HA and peak two is CS. Solid line, PH] counts; broken line,[35S] counts.

1979). Although the synthetic rates coincide closely, the change at 12-5 daysdoes not appear to be quite as sharp in the mutant limb bud as compared to thenormal control.

The increase in micrograms of CS over developmental time is shown inFig. 2D. The quantitative levels in the limb buds at the different ages are con-sistent with the integral of the rate function, as expected. In other words, thelinear increase in micrograms (Fig. 2D) corresponds to the constant segmentin the rate (Fig. 2A) while a change to quadratic increments (Fig. 2D) after12-5 days of development corresponds to the linear portion of the rate functionin Fig. 2A.* When this regression line is used to calculate x values (time indays) for the actual y values for + / + (micrograms of CS), the accumulationof CS in the brachypod limbs is observed to lag significantly (P < 0-01) behindtheir normal counterparts for each point past day 12 (paired t test for last fourtime points in Fig. 2D). The lag amounts to nearly one-half day by day 12-5through 14-5 (mean ± S.D. = 10 h ± 0-5 h).

* A regression line can be fitted to these data by weighting the time points that correspondto the quadratic increase by coefficients which give a linear transformation. Thus, when thetime points of days 1.1-5 through 14-5 are weighted by the following coefficients: — 2, —1,0,.1, 4, 9, 16, the resulting linear regression line produces a correlation coefficient, r2 = 0-99for + / + points, n = 8.

Glycosaminoglycan levels in normal and brachypod limbs 231

100 r

50

25

250

200

150en

1 iooXI'

6 50

40

o- 30x

2G

60

2P 40

20

15

10

20

15

10

12 13 14 12Gestational age (days)

13 14 15

Fig. 2. Incorporation of [3H]glucosamine and uronic acid concentration of macro-molecules from different-aged mouse limb buds. (A, D) Chondroitin sulfate; (B,E) hyaluronic acid; (C) glycoproteins; (F) units of hyaluronidase activity fromdifferent-aged mouse limb buds. Vertical bars in (E) and (F) represent 95% confi-dence limit of the mean {x±t 0-975 times S.E.). In (E) 4-8 determinations werecarried out for each point by the uronic acid assay and 25 determinations weremade with the Reissig test after Streptomyces HAdase digestion. In (F) threedeterminations were made for each point. Assay variances for the incorporation of[3H]glucosamine, uronic acid assay, and the Reissig test were all within 10%of the mean. Recovery of known quantities of HA and CS by the methods describedwas 90 % with a variance of 3 % within the mean, x — x, + / + ; O—O,bpH/bpyt.

Although the synthetic rates in the organ culture system coincide closely(Fig. 2A), differences were observed in the incorporation of [3H]glucosamineby 12-5-day limb mesenchyme grown in cell cultures (Table 3). Under theseculture conditions the normal cells had a significantly higher incorporation ratethan the mutant cells (alpha less than 0-05, two-sided t test, n = 4). The clearer

232 J. SHAMBAUGH AND W. A. ELMER

Table 3. Incorporation of [*H]glucosamine by 12-5-day normal andbrachypod limb mesenchyme in culture (sp. act. cpm/fig DNA)*

Genotype

Product + / + bpH/bpH

Glycoproteins 14-01 ±2-16f 17-46±6-69Hyaluronic acid 42-16± 11-38$ 29-48 + 7-39Chondroitin sulfate 28-96 ±4-39§ 20-85 ±2-98

* /^gDNA: + / + , 20-23±3-78; bpH/bpK, 23 08±600.t Mean ± S.D.X Statistically significant at a. < 010.§ Statistically significant at a < 005.

Table 4

Gestational

. Incorporationlimb

age (days)

of ^S]sulfate by normal and brachypodbuds in organ culture

Genotype

+ / + bpH/bpK

11-5-12-0 31212±7204* 27094±148512-25-12-75 23 868 ±6472 24145 ±301213-0-140 80383 ±15379 7O33O±29363

* Mean ± S.D., n = 3. Data are expressed as cpm/30 limb buds.

Table 5. Incorporation of [35S]sulfate by 12-5-day normal andbrachypod limb mesenchyme in culture

Genotype cpm/mg DNA

+ / + 4990 ±830*bp^/bp* 3380 ±370

* Mean±S.D., n = 4. Difference in means is significant at P < 005, 2-sided t test.

expression of a difference at 12-5 days could be related to the greater availabilityof the radioactive precursor to cells in a monolayer as opposed to an organculture.

The radioactive precursor, [35S]sulfate, was also used to determine sulfateincorporation into CS for normal and brachypod limb buds. It is apparent fromthe data in Table 4 that the variance in the incorporation of [35S]sulfate in organculture is too large to determine if any significant difference exists. However, asobserved with the incorporation of [3H]glucosamine (Table 3), the incorpora-tion of [35S]sulfate by 12-5-day limb mesenchyme grown in cell culture wassignificantly lower in the mutant as compared to controls (Table 5). The lower

Glycosaminoglycan levels in normal and brachypod limbs 233

incorporation of both [3H]glucosamine and [35S]sulfate into CS by brachypodcell cultures suggests that a decrease in synthesis is occurring rather than adecrease in sulfation of individual CS molecules.

Hyaluronic acid measurements

The incorporation for HA increases until day 12-5 of gestation and thendeclines (Fig. 2B). After day 12-5 the radioactivity is significantly higher in the+ / + limb bud as compared to bpn/bpa (alpha less than 0-01, two-sided pairedt test for last four time points; mean±s.D. of + / + minus bp^/bp11 = 5592±114 cpm, n = 6). On the other hand, the micrograms of HA at and before 12-5days is greater in the bp11 but returns to normal levels after day 12-5 (Fig. 2E;alpha less than 0-05, two-sided paired / test for time points 12 and 12-5; mean ±S.D. of + / + minus bp^fbp^ = -9-75 ±6-07, n = 4). Thus the levels of HAin the mutant limb bud are greater than expected from the incorporationrate over the entire time period tested and in fact are actually in excess beforeday 12-5. In addition, the levels of HA in both genotypes do not reflect theexpected integral of the rate function as was true for CS. Instead, a pronounceddampening of HA levels occurs during days 12-13 of gestation (Fig. 2E).

It has been reported that prior to the onset of overt chondrogenesis in thechick limb there is a decreased incorporation of labeled precursor into HAconcomitant with an increase in hyaluronidase (HAdase) activity (Toole,1972). In view of these observations, experiments were carried out to determinewhether similar changes in enzyme activity were also occurring in the mouselimb system. As seen in Fig. 2F there is a sharp increase in HAdase activitybetween days 12 and 13. The magnitude of the increase is the same for bothgenotypes, however, the two curves are not exactly coincident. The delay in theonset of HAdase activity in the bp^/bp11 limb buds could contribute to thehigher amounts of HA observed before 12-5 days of gestation, even thoughno statistical difference was demonstrated.

A series of experiments was carried out, therefore, to determine the turn-over time of HA in whole limb buds which were first cultured in the presenceof [3H]glucosamine for 6 h, then washed and further cultured in non-radioactivemedium containing an excess of unlabeled glucosamine (Fig. 3). The rate ofdegradation of HA in the normal limb bud shows only a slight change up to 2days of incubation, but increases sharply after the second day. When the slopesfor the entire lines of both normal and brachypod are calculated, the slope forbp^/bp3- is less negative (alpha less than 0-001) than for + / + , indicating aslower loss of radioactive GAGs. The total turnover time was determined to bebe 12 days for + / + and 13 days for bp

Equivalency of glycoprotein measurements

An explanation other than turnover for the differences seen between genotypesin HA levels and incorporation rates is that the precursor pool of UDP-iV-

234 J. SHAMBAUGH AND W. A. ELMER

2Days

Fig. 3. Turnover of glycosaminoglycans in organ culture. Twelve-day limb buds werelabeled at t0 with 5 /iCi/ml [3H]glucosamine for 6 h, followed by 4 days of incu-bation in the absence of tritiated precursor with a 50-fold excess of glucosamine.Vertical bars represent 95% confidence limit of the mean (x±t 0-975 times S.E.),n = four determinations for each point from three separate experiments. Slopes aresignificantly different P < 0-001, 2-sided / test for regression in two populations.x — x , + / + ; O - O , bpv/bp*.

acetylglucosamine is larger in the mutant and thus contains relatively less radio-active glucosamine than the control. Thus, the incorporation rates when com-pared to + / + would appear less than expected based on the actual levels ofHA at all ages tested. An excessively large precursor pool might also be relatedto the occurrence of excess HA in the mutant. However, arguing against thispossibility is the incorporation of [3H]glucosamine into GPs (Fig. 2C). AlthoughGPs utilize the same precursor as HA, the radioactivity from mutant and normalare not different from day 12-5 on (mean±s.D. of + / + minus bp^/bp11 =62 cpm, paired t test, n = 9; a at 0-05). In addition, there is no correlation overdevelopmental time between either [3H]HA and [3H]GP (r2 = 0-30) or [3H]CSand [3H]GP (r2 = 0-02, n = 9).

DISCUSSION

The low level of CS synthesis at day 11 of gestation corresponds to the timeduring which the non-specific CS proteoglycan is found in the mouse limb, whilethe sharp increase in synthetic rate at day 12-5 corresponds to the first appear-

Glycosaminoglycan levels in normal and brachypod limbs 235ance of the cartilage-specfic molecule (Royal & Goetinck, 1977). The delay inaccumulation of CS in the mutant supports previous observations made at theultrastructural level (Duke & Elmer, 1979). The decreased CS might be relatedto the excess HA found before day 12-5 of gestation. It has been shown thatHA is inhibitory to CS synthesis (Wiebkin & Muir, 1975, 1977; Handley &Lowther, 1976). Thus, the greater level of HA found in the mutant should actto suppress chondrogenesis. Alternatively, excess HA may block cell surfaceinteractions thought to be important to the initiation of chondrogenesis (Toole,1972; Levitt & Dorfman, 1974). Such a physical barrier between mutant cellsmay account for the observed delay in rearrangement of surface molecules thatcorrelates with differentiation (Hewitt & Elmer, 1978). A third possible actionof excess HA in the mutant involves the observation that cell adhesion is greateramong the mutant cells (Duke & Elmer, 1977). This aberrant adhesion mayaccount for the abnormal mesenchymal cell condensations observed both invivo (Griineberg & Lee, 1973) and in vitro (Elmer & Selleck, 1975) that occurjust prior to chondrification. Since it has recently been shown that HA is animportant factor in SV 3T3 cell adhesion (Underhill & Dorfman, 1978), excessHA in the mutant might be responsible for the increased adhesion of bpli cells.Thus, the exact manner in which HA is affecting chondrogenesis is unclear atpresent and could be multi-fold.

The relationship of HA and HAdase activity to cytodifferentiation is stilluncertain. Some reports suggest that HA is inhibitory to chondrogenesis(Toole, 19736; Solursh, Vaerewyck & Reiter, 1974), whereas, others indicatethat it has no effect (Finch, Parker & Walton, 1978). In the present study nolarge differences were found between genotypes in the magnitude of HAdaseactivity. There was, however, a delay in the appearance of activity, and a slowerturnover of labeled HA was observed for bp^/bp11. Excess GAG and decreasedturnover was described for this mutant in the neonate also (Rhodes & Elmer,1975). The loss of radioactivity as measured under the described conditions isnot great enough to account for either the differences in the incorporation ratesbetween the two genotypes at 12-5 days of age or the dampening of HA levelsobserved between days 12-13. One explanation for this is that degradation doesnot occur in culture at the same rate as it does in vivo. It should also be notedthat the presence of at least two metabolic pools for GAGs that turn over atdifferent rates has been previously described (Gross, Matthews & Dorfman,1960; Lohmander, 1977). Differential labeling of such pools may give an errone-ous impression of the magnitude of the turnover rate. However, based on thedata presented decreased turnover is still the most probable explanation forexcess HA accumulation in the mutant.

The onset of decreasing HA synthesis has the same inflection point as doesincreasing CS synthesis, again indicating the opposing roles of these two mole-cules. The pronounced dip in the HA synthetic rate at day 13 corresponds to thepeak of HAdase activity. It is noteworthy that at the same time, CS appears

236 J. SHAMBAUGH AND W. A. ELMER

totally unaffected by the increase in HAdase, even though other known mamma-lian HAdases attack both CS and HA (Saito, Yamagata & Suzuki, 1968;Hayashi, 1978). A similar specificity was reported in regenerating newt limb(Toole & Gross, 1971).

The regulatory mechanisms of HA levels that may be defective in brachypodare unknown at present. The excess HA in the mutant is not due to undersul-fated or unsulfated CS since (1) measurements of HA levels made with Strep-tomyces HAdase, which is specific for HA, also demonstrated excess HA inthe mutant, (2) the HA peak from DEAE chromatography was equally suscep-tible to leech-head HAdase for each genotype at the stage of excess HA in themutant, and (3) the 35S/3H ratio was similar for both genotypes. Thus, thismutation is apparently different from brachymorph, which displays a similarlimb phenotype but contains undersulfated CS (Orkin, Pratt & Martin, 1976).In addition, the observations that CS metabolism is nearly normal except for adelay in the onset of accumulation while HA levels are high indicate that bp11

is unlike other cartilage mutants that display reduced skeletal elements. Forexample, nanomelic chicks synthesize no cartilage-specific CS proteoglycan(Mathews, 1967; Pennypacker & Goetinck, 1976), while the hydrocephalic(ch+/ch+) mouse shows reduced synthesis of all GAGs (Breen, Richardson,Bondareff & Weinstein, 1973). In the achondroplastic (en/en) mouse, matrixmaterials are normal (Kleinman, Pennypacker & Brown, 1977), and thus theanomolous cell arrangements may account for the defective cartilage. In thechondrodysplasia mutation (cho/cho) of the mouse, abnormal arrangement ofcells is thought to be due to abnormal matrix (Seegmiller, Eraser & Sheldon1971). Brachypod may be related to the mouse mutation, amputated (Flint,1977) in which the entire skeleton is affected. The abnormal cartilage is thoughtto arise from increased cell adhesions that lead to defective migration of cellsinto pre-cartilage condensations. A similar conclusion could be drawn forbrachypod. The abnormal HA levels might increase cell adhesion and interferewith cell interactions of migration, and thereby lead to the incorrect or deficientpartitioning of cells into the mesenchymal condensations. In addition, the resi-dual HA may also suppress the chondrogenic program, which consequentlyresults in the formation of an abnormal limb.

REFERENCES

BITTER, T. & MUIR, H. M. (1962). A modified uronic acid carbazole reaction. Analyt. Bio-chem. 4, 330-334.

BREEN, M., RICHARDSON, R., BONDAREFF, W. & WEINSTEIN, H. G. (1973). Acidic glycosamino-glycans in developing sterno-costal cartilage of the hydrocephalic (ch+/ch+) mouse.Biochim. biophys. Acta 304, 828-836.

CORSIN, J. (1975). Influence du hyaluronate et de la hyaluronidase sur la chondrogenesecephalique chez les amphibiens. Acta Embryologiae Experimentalis 1, 15-22.

DELUCA, S., HEINEGARD, D., HASCALL, V. C, KIMURA, J. H. & CAPLAN, A. I. (1977). Chemi-cal and physical changes in proteoglycans during development of chick limb bud chondro-cytes grown in vitro. J. biol. Chem. 252, 6600-6608.

Glycosaminoglycan levels in normal and brachypod limbs 237DERBY, M. A. (1978). Analysis of glycosaminoglycans within the extracellular environments

encountered by migrating neural crest cells. Devi Biol. 66, 321-336.DUKE, J. & ELMER, W. A. (1977). Effect of the brachypod mutation on cell adhesion and

chondrogenesis in aggregates of mouse limb mesenchyme. / . Embryol. exp. Morph. 42,209-217.

DUKE, J. & ELMER, W. A. (1978). Cell adhesion and chondrogenesis in brachypod mouselimb mesenchyme: fragment fusion studies. / . Embryol. exp. Morph. 48,161-168.

DUKE, J. & ELMER, W. A. (1979). Effect of a brachypod mutation on early stages of chondro-genesis in mouse embryonic hind limbs: an ultrastructural analysis. Teratology 19, 367-375.

ELMER, W. A. (1976). Morphological and biochemical modifications of cartilage differentia-tion in the brachypod and other micromelic mouse embryos. In Mechanisms de la Rudi-mentation des Organes chez les Embryons de Vertebres (ed. A. Raynaud), pp. 235-242.Paris: Centre National de la Researche Scientifique.

ELMER, W. A. & SELLECK, D. K. (1975). In vitro chondrogenesis of limb mesoderm from nor-mal and brachypod mouse embryos. / . Embryol. exp. Morph. 33, 371-386.

FINCH, R. A., PARKER, C. L. & WALTON, S. T. (1978). The lack of an inhibitory effect ofhyaluronate on chondrogenesis in chick limb-bud mesoderm cells grown in culture. CellDifferentiation 7, 283-293.

FLINT, O. P. (1977). Cell interactions in the developing axial skeleton in normal and mutantmouse embryos. In Vertebrate Limb and Somite Morphogenesis (ed. D. A. Ede, J. R. Hinch-liffe & M. Balls), pp. 465-484. Cambridge University Press.

GILES, K. & MYFRS, A. (1965). An improved diphenylamine method for the estimation ofdeoxyribonucleic acid. Nature, Lond. 206, 93.

GOETINCK, P. R., PENNYPACKER, J. P. & ROYAL, P. D. (1974). Proteochondroitin sulfatesynthesis and chondrogenic expression. Exp I Cell Res. 87, 241-248.

GREENBERG, J. H. & PRATT, R. M. (1977). Glycosaminoglycan and glycoprotein synthesisby cranial neural crest cells in vitro. Cell Differentiation 6, 119-132.

GROSS, J. I., MATTHEWS, M. B. & DORFMAN, A. (1960). Sodium chondroitin sulfate-proteincomplexes of cartilage. / . biol. Chem. 235, 2889-2892.

GRUNEBERG, H. & LEE, A. J. (1973). The anatomy and development of brachypodism in themouse. / . Embryol. exp. Morph. 30, 119-141.

HANDLEY, C. J. & LOWTHER, D. A. (1976). Inhibition of proteoglycan biosynthesis by hya-luronic acid in chondrocytes in cell culture. Biochim. biophys. Acta 444, 69-74.

HASCALL, V. C, OEGEMA, T. R., BROWN, M. & CAPLAN, A. I. (1976). Isolation and charac-terization of proteoglycans from chick limb bud chondrocytes grown in vitro. J. biol.Chem. 251, 3511-3519.

HAT\E, Y. & MAKITA, A. (1975). Colorimetric determination of hyaluronate degraded byStreptomyces hyaluronidase. Analyt Biochem. 64, 30-36.

HAYASHI, S. (1978). Study on the degradation of glycosaminoglvcans by canine liver lysosomalenzymes. / . Biochem. 83, 149-157.

HEWITT, A. T. & ELMER, W. A. (1978). Developmental modulation of lectin-binding sites onthe surface membranes of normal and brachypod mouse limb mesenchyme cells. Differen-tiation 10, 31-38.

HOLTZER, H., OKAYAMA, M., BIEHL, J. & HOLTZER, S. (1978). Chondrogenesis in chicklimb buds and somites. Experientia 34, 281-284.

KLEINMAN, H. K., PENNYPACKER, J. P. & BROWN, K. S. (1977). Proteoglycan and collagen of'achondroplastic' (cn/cn) neonatal mouse cartilage. Growth 41, 171-177.

KOCHHAR, D. M. & AYDELOTTE, M. B. (1974). Susceptible stages of abnormal morphogenesisin the developing mouse limb analyzed in organ culture after transplacental exposure tovitamin A. / . Embryol. exp. Morph. 31, 721-734.

KVJST, T. N. & FINNEGAN, C. V. (1970). The distribution of glycosaminoglycans in the axialregion of the developing chick embryo. II. Biochemical analysis. / . exp. Zool. 175, 241-258.

LEVITT, D. & DORFMAN, A. (1974). Concepts and mechanisms of cartilage differentiation.Current Topics in Developmental Biology (ed. A. Moscona & A. Monroy) 8, 103-149.

l6 EMB 56

238 J. SHAMBAUGH AND W. A. ELMER

LINKER, A. & HOVINGH, P. (1972). Heparinase and heparitinase from Flavo-bacteria. Meth.Enzymol. 28, 902-911.

LOHMANDER, S. (1977). Turnover of proteoglycans in guinea pig costal cartilage. ArchsBiochem. Biophys. 180, 93-101.

MATTHEWS, M. B. (1967). Chondroitin sulfate and collagen in inherited skeletal defects ofchickens. Nature, Lond. 213, 1255-1256.

ORKIN, R. W., PRATT, R. M. & MARTIN, G. R. (1976). Undersulfated chondroitin sulfate inthe cartilage matrix of brachymorphic mice. Devi Biol. 50, 82-94.

ORKIN, R. W. & TOOLE, B. P. (1978). Hyaluronidase activity and hyaluronate content of thedeveloping chick embryo heart. Devi Biol. 66, 308-320.

PENNYPACKER, J. & GOETINCK, P. F. (1976). Biochemical and ultrastructural studies of colla-gen and proteochondroitin sulfate in normal and nanomelic cartilage. Devi Biol. 50,35-47.

POLANSKY, J. R., TOOLE, B. P. & GROSS, J. (1973). Brain hyaluronidase: changes in activityduring chick development. Science 183, 862-864.

PRATT, R. M., GOGGINS, J. F., WILK, A. L. & KING, C. T. G. (1973). Acid mucopolysaccba-ride synthesis in the secondary palate of the developing rat at the time of rotation and fusion.Devi Biol. 32, 230-237.

PRATT, R. M., LARSEN, M. A. & JOHNSTON, M. S. (1975). Migration of cranial neural crestcells in a cell-free hyaluronate-rich matrix. Devi Biol. 44, 298-305.

REISSIG, J. L., STROMINGER, J. L. & LELOIR, L. F. (1955). A modified colorimetric method forestimation of N-acetylamino sugars. / . biol. Chem. Ill, 959-966.

RHODES, R. K. & ELMER, W. A. (1975). Aberrant metabolism of matrix components in neo-natal fibular cartilage of brachypod (bpH) mice. Devi Biol. 46, 14-27.

ROYAL, P. D. & GOETINCK, P. F. (1977). In vitro chondrogenesis in mouse limb mesenchymalcells: changes in ultrastructure and proteoglycan synthesis. / . Embrvol. exp. Morph. 39,79-95.

SAARNI, H. & TAMMI, M. (1977). A rapid method for separation and assay of radiolabeledmucopolysaccharides from cell culture medium. Analyt. Biochem. 81, 40-46.

SAITO, H., YAMAGATA, T. & SUZUKI, S. (1968). Enzymatic methods for the determination ofsmall quantities of isomeric chondroitin sulfates. / . biol. Chem. 243,1536-1542.

SEEGMILLER, R., FRASER, P. C, & SHELDON, H. (1971). A new chondrodystrophic mutantin mice. Electron microscopy of normal and abnormal chondrogenesis. / . Cell Biol. 48,580-593.

SMITH, G. N., JR., TOOLE, B. P. & GROSS, J. (1975). Hyaluronidase activity and glycosamino-glycan synthesis in the amputated newt limb: comparison of denervated, nonregeneratinglimbs with the regenerates. Devi Biol. 43, 221-232.

SOLURSH, M. (1976). Glycosaminoglycan synthesis in the chick gastrula. Devi Biol. 50,525-530.

SOLURSH, M. & MORRISS, G. M. (1977). Glycosaminoglycan synthesis in rat embryos duringthe formation of the primary mesenchyme and neural folds. Devi Biol. 57, 75-86.

SOLURSH, M., VAEREWYCK, S. A. & REITER, R. S. (1974). Depression by hyaluronic acid ofglycosaminoglycan synthesis by cultured chick embryo chondrocytes. Devi Biol. 41,233-244.

TOOLE, B. P. (1972). Hyaluronate turnover during chondrogenesis in the developing chicklimb and axial skeleton. Devi Biol. 29, 321-329.

TOOLE, B. P. (1973a). Hyaluronate and hyaluronidase in morphogenesis and differentiation.Am. Zool. 13, 1061-1065.

TOOLE, B. P. (19736). Hyaluronate inhibition of chondrogenesis: antagonsim of thyroxine,growth hormone, and calcitonin. Science, N.Y. 180, 302-303.

TOOLE, B. P. & GROSS, J. (1971). The extracellular matrix of the regenerating newt limb:synthesis and removal of hyaluronate prior to differentiation. Devi Biol. 25, 57-77.

UNDERHILL, C. & DORFMAN, A. (1978). The role of hyaluronic acid in intercellular adhesionof cultured mouse cells. Expl Cell Res. Ill, 155-164.

WIEBKIN, O. W. & MUIR, H. (1975). Influence of the cells on the pericellular environment.Phil. Trans. R. Soc. Lond. B 271, 283-291.

WIEBKIN, O. W. & MUIR, H. (1977). Synthesis of proteoglycans by suspension and monolayerculture of adult chondrocytes and de novo cartilage nodules - the effect of hyaluronic acid./ . CellSci. 27, 199-211.

(Received 18 July 1979, revised 2 October 1979)