Segment-long-spacing crystallites and reconstituted fibrils have been made from collagen ex- tracted from the ejaculatory duct of the adult male locust, Locusta ...
Eur. J. Biochem. 103, 75-83 (1980)
Locust Collagen : Morphological and Biochemical Characterization Doreen E. ASHHURST and Allen J. BAILEY Department of Anatomy, University of Birmingham, and Agricultural Research Council Meat Research Institute. Langford (Received July 2, 1979)
Segment-long-spacing crystallites and reconstituted fibrils have been made from collagen extracted from the ejaculatory duct of the adult male locust, Locusta migratoria. These show the same banding pattern after positive staining as segment-long-spacing crystallites and fibrils made from mammalian type I collagen. Native fibrils show the same periodic pattern as type I fibrils, but it is not so distinct. Biochemical analysis of pepsin-digested locust collagen shows that there are two collagenous components. The a chains of the major component are similar to mammalian a1 (I) chains, except that the number of hydroxylysine residues is elevated and the CNBr peptides differ. There are no a2 chains; hence this locust collagen molecule is an a1 trimer. The second component, which is present in only minute quantities, may be a type IV basement membrane collagen. It is concluded that the fibrous collagen molecule of the locust is very similar to that of mammalian type I trimers and to those of other invertebrates which have been examined.
Since the chitinous exoskeleton serves the main supporting function in an insect, it was assumed that insects possess little, or no, collagen [l]. This is now known to be untrue, since collagen in a fibrous form, or as basement membrane, is found throughout the insect where it provides supporting tissues around the organs and under the epidermis [2]. Insect collagen fibrils are very variable; those in flies and moths are of very small diameter and indistinctly banded [3,4], while in other insects, such as locusts and cockroaches, the fibrils are of much larger diameter and display a distinct axial periodicity in sections of about 62 nm [5]. The collagenous nature of the fibrils was originally established by amino-acid analyses which indicated the presence of hydroxyproline in acid hydrolysates [6-81. More rigorous studies of insect collagens have been hampered by the very small amounts of collagen present in any one site in the insect. Studies of other invertebrate collagens have revealed some striking similarities in these and vertebrate collagens. Nordwig et al. [9] made segment-long-spacing crystallites and reconstituted fibrils from extracts of the sea anemone, Actinia equina, the liver fluke, Fasciola hepatica and the snail, Helixpomatia, which were similar in all respects to segment-long-spacing crystallites and reconstituted fibrils made from calf skin, type I, collagen. The main difference in the amino-acid analyses of Enzyme. Pepsin (EC 3.4.23.1).
these collagens when compared to calf skin collagen is the large number of hydroxylysine residues. Further studies of the sea anemone collagen have shown that the molecule has three identical CI chains and that the chains are cleaved into 12 cyanogen bromide (CNBr) peptides [lo], which contrasts with the two a1 and one a2 chains of the mammalian type I molecule. As yet, no other invertebrate collagen which forms banded fibrils has been studied extensively. The collagens of the annelid and nematode cuticles and the byssus threads of molluscs do not form typically banded collagen fibrils and will not be considered further in this paper. The similarities of the collagen molecules of the invertebrates with those of mammalian type I collagen are an interesting reflection on the evolution of this ubiquitous structural protein. It seemed of interest to investigate the insect collagen molecule further since this had been made more feasible by the discovery of a thick layer of collagen around the ejaculatory duct of the adult male locust, Locusta migratoria [ll]. The collagen fibrils in this layer of connective are produced mainly in the first six days after the adult tissue moult [5] by fibroblasts which closely resemble those of mammalian tissues and which synthesize the collagen via a similar secretory cycle [12]. The ease with which locusts can be bred in large numbers and the duct dissected out makes this source of insect collagen suitable for experimental work. A series of experiments
16
were designed to characterize the locust collagen molecule both by the electron microscopical observation of reconstituted fibrils and by biochemical analyses. MATERIALS AND METHODS
Locust Collagen
left overnight at 4 "C. The homogenate was centrifuged at 2500 rev./min and the pellet was resuspended in 0.5 ml of the EDTA solution. One drop of the suspension was placed on a collodion-carbon-coated grid. Several drops of 1 % phosphotungstic acid, pH 7.4, were dropped onto the grid before it was drained and dried. This staining was done at room temperature.
ELECTRON MICROSCOPICAL STUDIES
Preparation of Soluble Collagen
Sectioned Fibrils
Locust collagen is very insoluble and so in order to obtain soluble collagen, newly moulted adult male locusts, Locusta migratoria, were fed on wheat seedlings sprayed with a 1 % aqueous solution of p-aminoproprionitrile. The experiment was repeated several times, but on average the cage contained about 70 locusts and over a period of 6 days about 100 ml of the p-aminoproprionitrile solution was sprayed onto the wheat, all of which was consumed. The locusts were maintained at a day temperature of about 34 "C and a night temperature of about 26 "C. After 6 days, the ducts were removed and homogenized in 2 m l 0.5 M NaC1. The homogenate was left at 4°C for 2- 3 days and then centrifuged at 45000 x g for 1.5 h. The supernatant was retained and the pellet was resuspended in 2 ml 0.5 M NaCl. This was repeated twice and then the pellet was resuspended in 2 ml 0.05 M acetic acid and extracted for 2- 3 days before centrifugation ; the acid extraction procedure was repeated once more.
The tissue was prepared for electron microscopy as described by Ashhurst and Costin [5]. Rat Tail Tendon Collagen
An extract of mammalian type I collagen from rat tail tendon was prepared in the same way as the locust collagen. Segment-long-spacing crystallites and reconstituted fibrils were made and stained as above for comparison with the locust preparations. Electron Microscopy
The crystallites and fibrils were examined with an A.E.I. EM 801 electron microscope operated at 80 kV. The magnification of the micrographs was estimated from the known periodicity of the fibrils, i.e. 67 nm, established by X-ray diffraction of native, hydrated, rat tail tendon collagen, or from the distance between bands 6 and 22 (a1 CNBr-8 peptide) of the segmentlong-spacing crystallite, i.e. 79 nm [13].
Preparation of Segment-Long-Spacing Crystallites
Two or three drops of the first acid extract were dialysed for 3 days against 50 ml 0.4% ATP (Sigma) in 0.05 M acetic acid at 4 "C. Preparation of Reconstituted Fibrils
Two or three drops of the first acid extract were dialyzed against 50 ml 0.02 M Na2HP04 for 3 days at 4 "C. Positive Staining of Crystallites and Fibrils
One drop of dialyzate containing the suspension of either crystallites or fibrils was placed on a collodion-carbon-coated grid and left for 10 min. The excess fluid was then drained off and the grid allowed to dry. The preparation was stained with 0.1 % aqueous phosphotungstic acid, pH 2.6 for 10 min, rinsed and dried, followed by a similar procedure with 0.1 % aqueous uranyl acetate, pH 4.5. All staining was done at 4 "C. Preparation of Negatively Stained Fibrils
About 10 ducts from 25-day adult locusts were homogenized in 0.5 ml 4 % Na-EDTA, pH 7.4, and
BIOCHEMICAL STUDIES
These were performed on ejaculatory ducts removed from 5 - 8-week-old adult male locusts. Analysis of Solubilized Collagen Pepsin Digestion of Locust Collagen. Approximately 650 ejaculatory ducts were washed in physiological saline (0.9 % NaCI, pH 7.4) overnight at 4 "C, centrifuged, extracted overnight in 1.0 M NaCl in 0.05 M Tris/HCl buffer, pH 7.5 and then transferred to 0.5 M acetic acid. Pepsin (Worthington, 3250 units/ mg) at an enzyme substrate ratio of 10: 1 was added to the suspension and the digestion continued for 24 h at 15 "C. The reaction was stopped by raising the pH to 8.0 for 10min and then lowering the pH to 4.0. The digest was then dialysed against 0.5 M acetic acid overnight, centrifuged, and stored frozen. Salt Fractionation. The solubilized collagen was purified by reprecipitation from the acetic acid solution with 5 % NaCl and then dissolved in 1 M NaC1, 0.05 M Tris pH 7.5. The collagen was fractionated by the stepwise addition of NaCl up to 4.0 M NaCl [14]. Precipitates formed at any stage were removed
77
D. E. Ashhurst and A. J. Bailey
by centrifugation and the procedure continued. The precipitates were dialysed against 0.5 M acetic acid and lyophilized prior to analysis by sodium dodecylsulfate/polyacrylamide gel electrophoresis, CM-cellulose chromatography and the amino acid analyzer. Further purification of the precipitates was achieved by the thermal gelation technique of Trelstad et al. [15]. Sodium DodecylsulfatelPolyacrylamide Gel Electrophoresis. The precipitate obtained from the salt fractionation procedure was redissolved and denatured in 2 sodium dodecylsulfate at 60 "C with, and without mercaptoethanol, to cleave any disulphide bonds present. The a-chain mobilities were compared against known collagen types using the flat-bed technique employing Tris/borate buffer at pH 8.5 as previously described [161. Amino-Acid Analysis. A portion of the lyophilized precipitate was hydrolysed in glass-distilled constantboiling 6 M HC1 for 24 h in sealed tubes at 105 "C. The HCI was removed by rotatory evaporation under vacuum and the residue analysed with a Jeol 6 AH DK automatic amino acid analyzer. Analysis of Cyanogen Bromide Peptides. A portion of the lyophilized precipitate was digested under nitrogen with cyanogen bromide (CNBr) in 70% formic acid at 30 'C for 4 h. The samples were diluted with deionized water and the excess CNBr and formic acid removed by evaporation in vucuo at 30°C. The residue was redissolved in 0.5 M acetic acid and centrifuged. The peptides were separated by electrophoresis on disc gels using phosphate buffer [17]. After staining with Coomassie blue the gels were analysed by comparison of the densitometric tracings using a Joyce-Loebl Chromoscan, with peptide patterns of known collagens. Identification o j Reducible Crosslinks Approximately 200 ducts were homogenized and washed in physiological saline (0.9 %, NaCI, p H 7.4) overnight at 4 ° C and then centrifuged. The ducts were resuspended in the physiological saline and reduced with tritiated potassium borohydride at a collagen : KB3H4 ratio of 30:l for 1 h at 15 "C; the reaction was stopped by the addition of acetic acid to lower the pH to 4. The suspension was dialysed against water, lyophilized and then hydrolysed with 6 M HCl. The reducible crosslinks were identified as previously described in detail [18]. Briefly, the amino acids in the hydrolysate were separated on ion-exchange resins using volatile pyridinelformate buffers, and the reduced crosslinks were identified by their tritium activity. Confirmation of the identity of these radioactive components was achieved by rechromatography of the separated components on the Jeol automatic
amino acid analyzer and comparison of their elution positions with authentic standards.
RESULTS The ejaculatory duct is part of the male reproductive system of the locust and it develops as an invagination of the epidermis. In the newly moulted adult the duct is surrounded by spindle-shaped cells which are differentiating into fibroblasts. Over a period of 10 days, they synthesize collagen and the associated glycosaminoglycans to produce a thick layer of fibrous tissue around the duct (Fig.1) [S]. In the sexually mature adult many of the collagen fibrils are very large and irregular in cross-section (Fig. 3) and show a clear axial periodicity (Fig. 2) [5]. ELECTRON MICROSCOPICAL STUDIES
Banding Pattern of the Fibrils The detail of the bands within each repeating period are most clearly seen in positively stained reconstituted fibrils (Fig. 4). If this fibril is compared with a similar fibril made from rat tail tendon, type I, collagen (Fig.4), it can be seen that the pattern of bands within each period is identical along the two collagen fibrils. Thus the bands along the positively stained locust fibrils may be given the same notation (I-XII) devised by Bruns and Gross [19] for vertebrate type I fibrils. The only differences in the banding are minor and in the electron density of some of the bands; for example, band 11 tends to be less densely stained in the locust than in the rat fibril. Native locust collagen fibrils in sections (Fig.3) do not show the banding pattern as clearly, but the pattern is nevertheless present. After negative staining with phosphotungstic acid, the native locust fibrils (Fig.5) show alternate dark and light bands within each period. These two bands are further transected by other less prominent bands. The appearance is almost identical with that of a negatively stained native rat tail tendon fibril [20]. Segment-Long-Spacing CrystalliteJ A typical segment-long-spacing crystallite made from locust collagen is shown in Fig.6 alongside a similarly prepared crystallite made from rat tail tendon collagen. The pattern and arrangement of the bands are similar along both crystallites. Also their lengths are similar; the length of the locust collagen crystallite (bands 0- 57) [21] may be calculated from the known length of the ctl CNBr-8 peptide (bands 6-22) of 79 nm [13]. The length is 296 nm, compared with the figure of 294.7 nm for a crystallite of calf skin collagen calculated by Bruns and Gross [21].
78
Locust Collagen
Fig. 1. ( A ) Dirrgrutn of (1 ircniswrse scctioir of /hi, c,jmultrtorj. duct oj rhc, adult nzok loc,u.tt. f B j Drtrii.ing of crrm ii~dituic~d, a t higher rmgnificution. ( I ) Elongated epidermal cells; (2) connective tissue of basal layer; ( 3 ) connective tissue with fibroblasts; (4) fat body cells Fig. 2. A .section uf the ,fibrous connective tissue showing longitudinally orientated fi'hrils of large diczmcter und c,learl>.r i r f k e d periodicity oj 62 inn ( I S measured on thc section. Magnification x 140000 Fig. 3 . A section of fibrous tissue showing the very irregular cross-sections of the large-diumeter ,fi:hrils. Magnification x 140000
BIOCHEMICAL STUDIES
Collagen Isotypes Pepsin digestion results in complete solubilization of the duct, hence the collagen isolated is representative of the whole tissue. Isolation of the collagen types by salt fractionation from the pepsin-solubilized ducts results in a minor precipitate at 2.4 M NaCI, while the major precipitate occurs at 4.0 M NaC1. Sodium dodecylsulfate/acrylamide gel analysis of the 2.4 M NaCl precipitate reveals a single band with the mobility of a1 type I, but after incubation with mercaptoethanol a second a chain of slightly greater mobility appears (Fig. 7). The 4.0 M NaCl precipitate produces a single band with the mobility of a1 type I although on overloading the gels a small amount of
the mercaptoethanol-sensitive band was found to be present (Fig.7). The relative proportion of the mercaptoethanol-sensitive 2 chain is increased by reprecipitation and thermal gelation (Fig. 7, tracks 5 and 6).
Amino-Acid Analysis The amino acid composition of the 4.0 M NaCl precipitate is very similar to that of type I collagen except for a high hydroxylysine content of 27 residues per 1000 (Table l), compared with 5-8 residues per 1000 found in mammalian type I collagen. The 2.4 1z/1 NaCl precipitate, which is a mixture of the major 4.0 M NaCl component and the 2.4 M NaCl mercaptoethanol-sensitive component, has a similar composition, but the hydroxylysine content is even higher,
79
D. E. Ashliurst and A. .I.Bailey
Fig. 4. Rc~~on.~fiti4teNI cdlugm ,fihri/J posifive/,v stuinrd wirh /'/?"s~hUtL11zfi.sti(. ucid cind wanjl acetate. (A) Fibril made frotn locust collagen ; (B) fibril made from rat tail tendon collagen. The notation of the banding is that used by Bruns and Gross [19]. Magnification x 254000 (approx.)
Fig. 5. An isolated native least collagen ,fibril negatively .sfained wilh phosphorunxstic cicid. Magnification x 239 000 (approx.) 1
51
1 57 Fig. 6. Si.gmcnt-long-spvpcic.irzg crjstallites positively smined with phosphotungstic ucid and uranyl acetare. (A) Segment-long-spacing crystallite made from rat tail tendon collagen. (B) Segment-long- spacing crystallite made from locust collagen. The numbering of the bands is taken from Bruns and Gross [21] and starts at the N-terminal end of the molecules. Magnification x 228000 (approx.)
46 residues per 1000. Since the component precipitated at 2.4 M NaCl could not be separated from the a chain precipitated at 4.0 M, the composition of the mercaptoethanol-sensitive component was calculated (Table 1); the relative proportion of the two components was estimated from densitometer tracings of the acrylamide gels. The calculated amino acid analysis of the 2.4-M-precipitated component bears a close resemblance to that of basement membrane type IV collagen : for example, the lysine/hydroxylysine ratio is high, and the hydroxyproline and leucine contents are high. Analysis qf' the Cynnogen Bromide Peptides The CNBr peptide pattern of the 4.0 M NaCl precipitate on sodium dodecylsulfate gels is shown in Fig. 8. It is completely different from type I collagen
80
Locust Collagen Table 1. Amino acid composition of comporientsprecipitated by 2.4 M NaCl arid 4.0 M NaCl from pepsin digests of locust ejuculutory duct The 2.4 M NaCl precipitate contains a mixture of the 4.0 M NaCl component plus the ‘minor component’ in a ratio of 1 : 1.3 (determined by densitometric trace of acrylamide gels) ; the composition of the ‘minor component’ was calculated using this ratio. n.d. = not determined Amino acid
Amount in
4.0 M NaCl precipitate
2.4 M NaCl precipitate
Calculated composition of minor component
98.7 67.4 23.9 27.7 83.4 94.9 311.3 40.2 21.9 n.d. 32.3 48.3 5.2 16.0 46.8 15.7 2.1 54.2
90.3 83.0 23.7 23.6 76.8 67.8 300.7 33.4 26.2 n.d. 42.8 56.2 11.6 20.5 64.6 16.9 0.17 55.1
residues! 1000
Fig. 7. Sodium dodccylsulfkte ipolyucryluniide-gel electrophosesis qf the fiuctions obtained by precipitation of the pepsin digest ($locust
ejuculutosy ducts with NuCl. Tracks 1,2: standard rl and ct2 chains from rat tail tendon collagen, without, and with, mercaptoethanol respectively. Tracks 3,4: total pepsin digest of locust ducts, without, and with, mercaptocthanol respectively; note appearance of minor component of higher mobility following mercaptoethanol treatment. Tracks 5,6: 2.4 M NaCl precipitate from pepsin digest of locust duct, purified by reprecipitation and thermal gelation, without, and with, mercaptoethanol respectively, note the increased proportion of mercaptoethanol-sensitive component. Tracts 7,8 : 4.0 M NaCl precipitate from pepsin digest of locust duct, without, and with, mercaptoethanol respectively
and a1 type I obtained from rat tail tendon. Further, comparison with the CNBr peptide pattern of porcine type I1 collagen, which also precipitates at 4.0 M NaCI, failed to reveal any similarity.
Reducible Crosslinks Analysis of the crosslink pattern revealed that borohydride-reduced locust collagen possesses the reduced crosslinks, dihydroxylysinonorleucine and hydroxylysinonorleucine, in a ratio of 2 : 1 (Fig. 9). The artifactual crosslink histidinohydroxymerodesmosine [22] is not present in this tissue.
DISCUSSION The results presented in this paper indicate that the locust possesses two types of collagen. The predominant type extracted from the ejaculatory duct is fibrous and shares both morphological and biochemical properties with the collagens of other animals. The second minor collagenous component was detected on sodium dodecylsulfate/polyacrylamide gel electrophoresis and further analysis shows that it bears
Hydroxyproline Aspartic acid Threonine Serine Glutdmic acid Proline Glycine Alanine Valine Methionine Isoleucine Leu cin e Tyrosine Phenylalanine Hydroxylysine Lysine Histidine Arginine
109.3 46.7 24.1 32.9 92.2 129.8 324.8 49.0 16.3 n.d. 18.5 38.2 1.6 10.4 24.1 14.1 4.6 53.1
a close similarity to basement membrane (type IV) collagen. The amount of this collagen is very small and it probably comes from the basement membrane of the epidermal cells, which is obliterated as the fibrous tissue develops, and the basement membranes of the fat body cells which lie on the outer surface of the fibrous layer. The segment-long-spacing crystallites made from acid-soluble locust collagen display an identical pattern of 57 bands to that seen along crystallites made from calf skin, type I, collagen. Crystallites made from types I1 and I11 mammalian collagens can be distinguished by several minor variations in the pattern of bands [23,24]. The other segment-long-spacing crystallites made from invertebrate collagens, that is, sea anemone, liver fluke and snail, are almost identical to mammalian type I crystallites [9]. This similarity of the segment-long-spacing crystallites implies that the distribution of the polar and non-polar amino acids along the length of the collagen molecules of these diverse animals is the same and that the molecules are of approximately the same length, that is about 300 nm long. It might appear anomalous that invertebrate collagens which are a1 trimers should produce both segment-long-spacing crystallites which
81
D. E. Ashhurst and A. J. Bailey
Fig. 8. Sodium dodecylsulfatelpolyacrylumide disc gel electrophoresis of peptides obtained ,following treatment of the collagen precipitates with cyanogen bromide. Track 1, type 1 collagen i.e. ( ~ l ) ~ rtrack 2 ; 2, a1 type 1 collagen chain: track 3, type I1 collagen; track 4, 4.0 M NaCl precipitate from pepsin digest of locust duct
Dihydroxylysini Norleucine
3.c
. .-c
E
c
C
-x 0
c ._ > ._ c
HydroxylysinoNorleucine
U
0
f
._ c ._ I= X
m
b
i l I
Leu Phe Tyr
300Hyl Lys
Elution volume ( m l )
Fig. 9. Elution profile of radioactive components in an acid hydrolysute o j locust collagen reduced with tritiated borohydride
display the same morphological features after staining as type I collagen composed of [a1 (1)]2a2, but trimers of either al(1) or a2 chains form segment-longspacing crystallites identical to those of normal type I collagen [25].
The periodic banding pattern of the native and reconstituted locust collagen fibrils has been analysed according to the nomenclature devised by Bruns and Gross [19]. The same 24 light and dark bands seen after positive staining of reconstituted rat tail tendon fibrils can be clearly distinguished along the reconstituted locust fibril, though there are some minor differences in the intensity of staining of some of the bands. This might reflect sinall differences in the distribution of the polar amino acids in some regions of the locust collagen molecule. An identical series of bands can also be seen in positively stained, native, sectioned fibrils, but they cannot be identified so readily. After negative staining, native fibrils show dark and light bands with interbands which are similar to those of rat tail tendon fibrils. Fibrils with the same periodic banding pattern have been described in a wide variety of insects including cockroaches, grasshoppers, water bugs and primitive wingless insects [2,8,26- 281. The estimated periodicity of these patterns varies widely from 50 nm to almost 70 nm, but this variation is almost certainly artifactual. The absence of distinct periodic banding along the thin fibrils of the Lepidoptera and Diptera [3,4] is probably due to the inability of the present staining methods to produce sufficient contrast in fibrils of small diameter; the same is true of some thin vertebrate fibrils, Using special staining techniques Locke and Huie [4] detected a periodicity of about 50 nm in the fibrils of Cupodes ethlius. While further work is necessary to characterize the collagen molecules of the Diptera and Lepidoptera, it seems very probable that they are similar to those of the other insects. Collagen fibrils from a wide variety of other invertebrates also display this same periodic banding pattern. These include native or reconstituted fibrils from Porifera [29,30], Coelenterates [31], Platyhelminths [9], Brachiopods [32], Annelids [33], Molluscs 191, Echinoderms [34], Arachnids [35] and Crustaceans [36]. Again, the figures given for the length of the repeating period vary widely in the different collagens, and great stress has been laid on these measurements as a criterion for the identification of a collagen fibril. The detailed similarity in the banding patterns of segment-long-spacing crystallites and fibrils from many different animals leads to the obvious conclusion that the length of the helical portion of all these collagen molecules and the charge distribution of the polar amino acids along them must be the same. Therefore, when the molecules are assembled into fibrils according to the modified quarter-stagger overlap model of Hodge and Petruska [37], the resulting banding patterns must have the same periodicity. It follows that any variation from the periodicity of 67 nm (determined by X-ray diffraction of native,
82
hydrated, rat tail tendon collagen) measured on electron micrographs must be due to the preparative procedures. Thus, it is the banding pattern, not the measured periodicity of this pattern, which is the morphological criterion for the identification of a collagen fibril. The major precipitate from the pepsin-digested collagen occurs at 4.0 M NaCI, in contrast to the anticipated point at which type I collagen precipitates (2.4 M NaCl). Type I trimer and type I1 collagen have been reported to precipitate at 4.0 M NaCl [38], but comparison of the CNBr peptide pattern of these and locust collagen on sodium dodecylsulfate/acrylamide gels demonstrates that the locust collagen possesses a different pattern. The only other invertebrate so far examined is the sea anemone which has 12 CNBr peptides [lo], so the two invertebrate collagens are unrelated on this criterion. Amino-acid analyses have been made of collagens from most invertebrate phyla; many of these are tabulated and discussed by Mathews [39]. In all these collagens about a third of the residues are glycine; proline and hydroxyproline are present, and hydroxylysine occurs in greater amounts than in vertebrate collagens. The locust collagen fits into this general pattern of amino-acid composition. Only a few invertebrate collagens have been investigated further. Collagens from a variety of animals including sea anemones, tapeworms, liver flukes, earthworms (body wall), crabs, various molluscs, including the octopus and squid, sea cucumbers and ascidians have been identified as a1 trimers [9,10,40, 411, but more recently, Kimura and Matsuura [14] claimed that a2 chains are present in the collagens of the crab, lobster, octopus and squid. In this context the work of Pucci-Minafra and co-workers is interesting; they have shown that the collagen extracted from the spicules of the embryo sea-urchin, Pauacentrotus lividus, is an ctl trimer [42], but that in the Aristotle’s lantern of the same species possesses both a1 and a2 chains [43]. a1 (I) trimers are also found in mammals and birds. Until recently, it seemed that they might be associated primarily with pathological tissues [44], or with cultured cells [45], but trimers have now been detected in normal human skin and embryonic chick tendon and calvaria [46,47]. It is noteworthy that these trimers also precipitate at 4.0 M NaCl and possess high hydroxylysine contents. The cross-linking system of the locust fibrils is based on the same mechanism as vertebrate collagens [48]. The major cross-link is hydroxylysino-5-ketonorleucine and the stability of this cross-link accounts for the insolubility of this collagen. This same crosslink is also the major one in sea anemone, sponge and sea cucumber collagens [49,50]. The striking similarities of this locust and other invertebrate fibrous collagen molecules to the mam-
Locust Collagen
malian type I trimers leads one to conclude that this collagen molecule already existed in the primitive coelenterates; during the evolution of the higher invertebrates and vertebrates it seems that only small conservative changes were necessary to fit this fibrous protein for its role in very diverse animals. We are grateful to the Centre for Overseas Pest Research for sending us their aged, adult male locusts to increase our supplies for the biochemical study. We should like to thank Mr Alan Murdoch and the other technicians in the Department of Anatomy for their patience in dissecting out the ducts from the locusts. We are also grateful to Mr T. J. Sims and Mr J. Wallington for their technical assistance.
REFERENCES 1. Rudall, K. M. (1955) Symp. Soc. Eup. B i d . 9, 49-71. 2. Ashhurst, D. E. (1968) Annu. Rev. E n f o n d . 13. 45-74, 3. de Biasi. S . & Pilotto. F. (1976) .I. Suhmicrosc..C ~ , t o /8. . 337345. 4. Locke. M. & Huie, P. (1972) Tissue & Cdl, 4 , 601-612. 5. Ashhurst, D. E. & Costin, N. M. (1974) Tissue & Cell, 6 , 279 - 300. 6. Ashhurst, D. E. (1959) Q. J . Microsc. Sci. 100, 401-412. 7. Ashhurst, D. E. & Richards, A. G . (1964) J . Morphol. 114, 237 - 246. 8. Harper, E., Seifter, S. & Scharrer, B. (1967) J . Cell Biol. 33, 385 - 394. 9. Nordwig, A,, Rogall, E. & Hayduk, U . (1970) in Chemistry and Molecular Biology of the Intercellular Matrix (Balazs, E. A., ed.) pp. 27-41, Academic Press, London and New York. 10. Nowack, H. s( Nordwig,A. (1974) Eur. J . Biochem. 45, 333342. 11. Martoja, R. & Bassot, J.-M. (1965) C.R. Acad. Sci. (Paris) 261, 2954-2957. 12. Ashhurst, D. E. & Costin, M. N. (1976) J . Cell Sci. 20, 377403. 13. Rauterberg, J. & Kuhn, K. (1971) Eur. J . Biochem. 19, 398407. 14. Bailey, A. J. & Sims, T. J. (1977) J . Sci. Food Agric. 28, 565 570. 15. Trelstad, R. L. & Lawley, K. R. (1977) Biochem. Biophj.s. Res. Commun. 76, 376- 384. 16. Sykes, B. C. & Bailey, A. J. (1971) Biochem. Biophys. Res. Commun. 43, 340 - 346. 17. Furthmayer, H. & Timpl, R. (1971) Anal. Biochem. 41, 510516. 18. Robins, S. P., Shimokomaki, M. & Bailey, A. J. (1973) Biochem. J . 131, 771 -780. 19. Bruns, R. R. & Gross, J. (1974) Biupolymers, 13, 931 -941. 20. Tromans, W. J., Horne, R. W., Gresham, G. A. & Bailey, A. J. (1963) Z. Zellforsch. Mikrosk. Anat. 58, 798-802. 21. Bruns, R. R. & Gross, J. (1973) Biochemistry, 12, 808-815. 22. Robins, S. P. & Bailey, A. J. (1973) Biochem. J . 136, 657-665. 23. Stark, M., Miller, E. J. & Kuhn, K. (1972) Eur. J . Biochem. 27, 192-196. 24. Trelstad, R. L., Kang, A. H., Igarashi, S. & Gross, J. (1970) Biochemistry, 9, 4993 -4998. 25. Tkocz, C. & Kuhn, K. (1969) Eur. J . Biochem. 7,454-462. 26. Baccetti, B. (1961) Redia, 46, 1-7. 27. Frdn$ois, J. (1972) Arch. Anat. Microsc. 61, 279-300. 28. Smith, D. S. & Treherne, J. E. (1963) Adv. Insect Physiol. I , 401 - 484. 29. Garrone, R., Huc, A. & Junqua, S. (1975) J . Ultrastruct. Res. 52, 261 -275. -
83
D. E. Ashhurst and A. J. Bailey 30. Ledger, P. W. (1974) Tissue & Cell, 6 , 385-389. 31. Nordwig, A. & Hayduk, U. (1969) J . Mol. Biol. 44, 161 -172. 32. Reed, C. G . & Cloney, R. A. (1977) Cell Tissue Res. 185, 17-42. 33. Baccetti, B. (1967) J . Cell Biol. 34, 885-891. 34. Baccetti, B. (1967) Monitore Zool. Ital. 1, 201 -206. 35. Baccetti, B. & Lazzeroni, G. (1969) Tissue & Cell, I , 417-424. 36. Baccetti, B. & Bigliardi, E. (1969) %. Zellforsch. Mikrosk. Anat. 99, 25 - 36. 37. Hodge, A. J. & Petruska, J. A. (1963) in Aspects of Protein Structure (Ramachandran, G . N., ed.) 289 - 300, Academic Press, London and New York. 38. Trelstad, R. L., Kang, A. H., Toole, B. P. & Gross, J. (1972) J . Biol. Chem. 247, 6469-6473. 39. Mathews, M. B. (1975) Connective Tissue, pp. 318, SpringerVerlag, Berlin, Heidelberg. 40. Pikkarainen, J., Rantanen, J., Vastamaki, M., Lampiaho, K., Kari, A. & Kulonen, E. (1968) Eur. J . Biochem. 4, 555-560.
41. Kimura, S. & Matsuura, F. (1974) J . Biochem. ( T o k y o ) 75, 1231- 1240. 42. Minafra, S . , Pucci-Minafra, I., Casano, C. 81 Gianguzza, F. (1 975) Boll. ZOO^. 42, 205 - 208. 43. Pucci-Minafra, I., Galante, R. & Minafra, S. (1978) J . Submicrosc. Cytol. 10, 53 - 63. 44. Moro, L. 81 Smith, B. D. (1977) Arch. Biochem. Biophys. 182, 33-41. 45. Mayne, R., Vail, M. S. 81 Miller, E. J. (1975) Proc. Nut1 Acad. Sci. U . S . A . 72, 4512-4515. 46. Uitto, J. (1979) Arch. Biocheni. Biophys. 192, 371-379. 47. Jimenez, S. A,. Bashey, R . I., Renditt, M. 81 Yankowski, R. (1977) Biocliem. Biophys. Res. Commun. 78, 1354- 1361. 48. Bailey, A. J., Robins, S. P. & Balian, G. (1974) Nature (Lond.) 251, 105-109. 49. Bailey, A. J. (1971) FEBS Lett. 18, 154-158. 50. Eyre, D. R. & Glimcher, M. J. (1973) Proc. Soc. Exp. Biol. Med. 144, 400-403.
D. E. Ashhurst, Department of Structural Biology, St-George’s Hospital Medical School, Cranmer Terrace, Tooting, London, SW17 ORE, Great Britain A . J. Bailey, Agricultural Research Council Meat Research Institute, Langford, Bristol, BS18 7DY, Great Britain
