Clinical Allergy, 1975, Volume 5, pages 241-254
Isolation and characterization of cat allergens
C. R. STOKESflrtf/ M. W. TURNER Department of Immunology, Institute of Child Health, 30 Guilford Street, London
Summary A crude (saline soluble) extract of cat skins capable of eliciting a strong positive prick skin test in cat sensitive individuals was fractionated on Sephadex G200. Active fractions were pooled and successively fractionated on isoelectric focusing gradients of pH 3 5 and pH 4-5. Allergenic activity was localized in two peaks with mean isoelectric points of 4-1 and 4-35 respectively. On immunoelectrophoresis the allergen with pi =4-35 was associated with a protein which was subsequently found to be immunologically indistinguishable from serum albumin and to have a molecular weight of 69,000 daltons. The allergen with pi =4-1 migrated in the aj-region on immunoelectrophoresis and had a mean molecular weight of 55.000 daltons. This allergen was isolated and analysed for amino acid and carbohydrate content. A combined extract of both allergens coupled to microcrystalline cellulose and used in a RAST procedure readily distinguished between two groups of individuals classified as skin lest positive and skin test negative lo cat allergen. Introduction
The importance of danders of mammalian origin in allergic disease has long been recognized. In 1864 Salter had suggested that cat dander may induce asthmatic reactions but it was Goodale (cited by Wodehouse (1917)') who first demonstrated, in cat sensitive individuals, positive skin tests to ethanol extracts of cat fur. Skin test studies on larger groups of asthmatic patients have confirmed these earlier reports; co*ke (1927) found that 19",; of 1000 asthmatics gave positive skin reactions to cat dander and Frankland (1967). in a study of 451 asthmatic patients, found that 25% reacted to animal extracts, with cat as a major allergen. Early attempts to fractionate extracts of cat fur (Wodehouse, 1917) involved the use of both strong acids and alkalis. These studies (Wodehouse. 1917; Walker. 1917) suggested that cat serum was allergenic in over half of the smalt number of cat sensitive subjects tested. However, further detailed characterization of the allergens was delayed by the absence of appropriate physico-chemical techniques. The recent Correspondence: C. R. Stokes, Department of Immunology, Institute of Child Health, 30 Guilford Street. London WCIN lEH. 241
242
C. R. Stokes and M. W. Tumer
availability of such techniques has stimulated fresh interest in the problems of allergen characterization and isoelectric focusing (Varga & Ceska. 1972). gel filtration (Holford-Strevens, 1973) and polyacrylamide gel electrophoresis (Ohman. Lowell & Bloch. 1973. 1974; Brandt. Ponterius & Yman. 1973) techniques have all been applied to the characterization of cat allergens. The present study describes the characterization of two major allergenic components from the dander of the domestic cat. One such allergen has been isolated in immunochemically pure form and characterized with respect to its molecular weight, isoelectric point and constituent amino acids and oligosaccharides. Materials and methods Extraction of crude allergen The skins (including furl of a group of laboratory cats were defatted in acetone (4'C) for 24 hr and then extracted in saline (0-85%) containing 0 05/,, sodium azide for 48 hr at 4 C. Residual fat was removed by ether extraction of (he saline extract. The aqueous solution was then dialysed against distilled water and lyophilized. Get filtration of saline extract 20 ml of the saline extract (15 mg.mP' w/v) were passed through a Millipore filter (pore size 0 22/O and applied to a calibrated column of Sephadex G2OO(5 x9l-2 cm). Upward elution with O-I M Tris-HCl-0 2 M NaCI-2mM EDTANa, containing 0-02"; sodium azide pH 7-6 was performed at 22 C using a pumping speed of 2K ml.hr"'. Fractions were pooled, dialysed against distilled water and freeze dried. Allergen activity was determined by prick skin tests in a cat sensitive subject using fractions reconstituted to a constant volume. Fractions with activity were pooled and the product dialysed and lyophilized. Isoeleciric focusing of allergenic [gel-fiUration) fraction Isoetectrlc focusing was performed at lB^C using the LKB 8100-10 system with a sucrose density gradient. The column was emptied by pumping at I ml.min"' with continuous monitoring of the etuate at 254 nm (LKB Uvicord I). Fractions were collected at I min intervals and the pH immediately read. One percent carrier ampholytes [pH ranges 3 10 and 3-6 (LKB 8141 and 8142)) were used initially in order to localize allergenic activity. Preparative experiments were then performed using 4% ampholytes pH 3-6. The allergen active region from the Sephadex G200 fractionation (120 mg) was focused and on emptying the column a region of intense skin test reactivity was identified. Fractions from this region were pooled and refocused in a second run. without the addition of any more ampholytes. by diluting one quarter of the pooled solution with distilled water to form the light (i.e. less dense) solution and diluting the remainder with 55"-,, sucrose solution (w/v) to form the dense solution. On completion of the second run the fractions were dialysed extensively against distilled water and finally adjusted to a constant volume. Prick skin testing Skin testing was performed on a cat-sensitive subject by piercing the derniis (21 gauge needle) through a droplet of allergen solution which had been previously sterilized by passage through a Millipore filter (0-22 fi). Tests were performed in
Cat allergens
243
duplicate on random numbered sites on the volar aspect of the forearm. Skin test solutions were diluted and used at increasing concentrations until a wheal and Hare were obtained. Wheal diameters were read in two different directions by the recipient using a negative saline control and a commercial cat allergen (Bencard) as positive control. The recipient read the tests using a number code for each sample. Imnninoelectroplwresis Immunoelectrophoresis was performed according to the method of Grabar & Williams (1953). using 1-5% (w/v) Oxoid lonagar in barbitone buffer pH 8-6 (I = 0-045). Immunodiffusion analyses Double diffusion was performed according to the method of Ouchterlony (1958) using 1-5% Difco Noble agar in phosphate butfer pH 8 0. Antisera Antisera to the initial saline extract (anti-CE) were raised in rabbits. Three mg of extract in complete Freund's adjuvant were injected into three sites and after 6 weeks a further 5-8 mg of alum precipitated extract were injected. A second injection of alum precipitated extract was given 8 weeks later and the animals were bled after a further 2 weeks. Apparent molecular weights The apparent molecular weights of the allergens were determined according to the method of Andrews (1964) from their elution volumes on a column of Sephadex G-200 using a bulTer of 0-1 M Tris-HCI-0-2 M NaCl-2mM EDTANa, containing 0 02'% sodium azide. pH 7-6. Calibration of the column was based on the elution volumes of the following proteins: human IgG (mol. wt 150.000), human serum albumin (mol. wt 69,000), ovalbumin (mol. wt 45,000), pFc' fragment of human IgG (Turner & Bennich. 1968) (mol. wt 26,700). Dissociation and reduction experiments The effect of dissociation on the apparent molecular weight of the cat allergen pl = 4-1 was studied by diaiysing a sample of the allergen against 0-1 M acetic acid 25 mM sodium chloride pH 2 2 followed by fractionation on a calibrated column of Sephadex G 200, equilibrated in the dissociating buffer. The apparent molecular weight of cat allergen pi =4-1 after reduction and alkylation was also obtained after gel filtration under dissociating conditions. Reduction was performed in de-aerated 0 55 M Tris-HCl-2mM EDTA Naz, pH 8 2 at 20°C under nitrogen. The sample was dialysed overnighi. de-aerated and saturated with nitrogen and 2-niercaptoethanol (Koch-Light Ltd) added to give a 0-1 M solution. After 1 hr the reaction was stopped by alkylation with a 10"^ moiar excess (over mercaptoethanol) of iodocetamide (Sigma Ltd) for 60 min at OC (ice water) in the dark. The sample was then applied to the calibrated column o^ Sephadex G-200 (see above). Amino acid composition A sample (1 mg) of the cat allergen pi 4-1 (see results) was taken for hydrolysis in 0-5 ml of 6 N hydrochloric acid, containing 7-05 ng of norleucinc (internal standard).
244
C. R, Stokes and M. W. Turner
Hydrolysis was performed in vacua at I IOC for 24 hr as described by Van Hofsten, Van Kley & Eaker (19651. The hydrolysate was lyophilized and redissolved in 0 25 ml of lithium citrate buffer pH 2 2 and analysed according to the method ofSpackman, Stein & Moore (1958). using a Beckman 121M amino acid analyser equipped with an integrator;computer (Autolab-System AA). Cysteic acid and tryptophan were not determined. Carbohydrate composition A sample (I mg) of cat allergen pi 41 was analysed for sugars using standard gas chromatographic procedures (Clamp, Dawson & Hough, 1967). RadioailergosorbetU tests The allergen rich material obtained after ge! filtration on Sephadex Ci-200 was coupled to micro-crystalline cellulose (Sigma-cell) and used for the measurement of specific IgE antibodies as described by Wide. Bennich & Johansson (1967). The amount of allergenic material required for coupling was titrated using a range of concentrations of the allergen and a constant amount of serum from a cat sensitive individual. For subsequent assays 0-10 mg of the allergen preparation was coupled to 1 mg of Sigmacell. For the performance of the test 0 5 ml of the allergen-Sigma cell suspension was mixed with 50/H of serum and incubated at room temperature for 16 hr with continuous mixing. The cellulose particles were then washed six times with physiological saline containing 1",, Tween-20and after the final wash lOO/d of ''^l-labelled specific rabbit anti-human IgF globulin (Fc,-immunosorbent purified) was added to each tube and the contents incubaled with agitation at room temperature overnight. The supernatant was then removed and the cellulose particles washed five times with physiological saline containing 1/,, Tween-20. Residual radioactivity was then counted in a Gamma counter. Kcsults Cat pelts were found to provide a ready source of crude allergen extract and at least 500 mg of lyophili/able material were obtained from each animal. A solution of crude allergen extract in physiological saline (15 mg.mL ') was found to be of comparable strength to the Bencard cat fur extract 150",, (Balch no. 3204 C0979). Immunoelectrophoretic analysis of the allergen extract using lhe rabbit anti-extract (anli-CH) revealed the presence of at least thirteen components. Fractionation of the saline extract on Sephadex G 200 gave reproducible elution profiles with large amounts of low molecular weight material (see Fig. 1). Prick skin tests on fractions from these separations localized allergenic activity to the molecular weight region 26.OCX) 68.000. Immunoelectrophoresis and double diffusion analysis of individual fractions within this region revealed the presence of multiple components in all samples tested. Further fractionation of the allergen rich material from gel-filtration separations was achieved by isoelectric focusing with I",, ampholytes on pH gradients 3 10 and 3-6. These experiments showed skin test activity in the pH region 4 5 (see Fig. 2). When preparative runs were performed using the double focusing schedule two peaks of activity were resolved (see Fig. 2 lower trace). The first peak was associated with fractions 48-62 having a mean pi of 4-1 and the second peak was associated with
Cat allergens
245
fractions 70-94, having a mean pi of 4-35. Double diffusion analysis (Fig. 3) of alternate fractions from the skin test active regions using the anti-cat extract antiserum Moleculor weight colibrotion
Fig. I. Gel filtration of saline extract (20 ml) of cat skin (15 mg.ml"' w/v) on a column of Sephadex G-200 (5x91 2 cm). The column was eluted by upward How using 01 M Tris-02 M NaCI-2niM EDTA Na2 containing 002",, NaNj, pH 7-6. Flow rate 2S ml.hf'. Temperature 22 C. Allergen activity, delermined by prick tests in a cat sensitive subject, was localized to the stippled region. Fractions in this region were pooled and lhe product dialysed and lyophilized.
showed the presence of a maior precipitin line in fractions 48-58 which was also detectable in fractions 60 and 62. The distribution and intensity of this precipitin line parallels closely the quantitative distribution of skin test activity in this region. Similarly a precipitin line Hrst detectable in fraction 70, strongly represented in fractions 74-84, and still detectable in fraction 94 appears to parallel the distribution of the second peak of skin test activity. Immunoelcctrophoretic analysis of alternate fractions around the first peak of skin test activity (pl = 4-l) (see Fig. 4) showed the presence of an anodic arc in the a-electrophoretic region (see arrow). Fractions 56-60 also contain an increasing amount of a trace component of faster mobility but the distribution of this protein did not parallel the distribution of skin test activity. Immunoelectrophoretic analysis of fractions corresponding to the second region of skin test activity (pi =4-35) are shown
246
C. R. Stokes and M. W. Turner
in Fig. 5. Ail fractions analysed contained at least three precipitating components atid the major arc (see arrow) showed a sitnilar distribution to the second region of skin test activity. This major component was further studied in gel tests and showed a reaction of complete identity with isolated cat albumin on double diffusion analysis.
5-0
5-0
•« E - E
70
90
110
150
Fraction no.
Fig. 2. Isoelectric loeusing profile of allergen-rich material (120 mg) from the gel-filtration experiment (Fig. I) using 4",, ampholytes and a pH range 3 0-6 0 (upper trace). Allergen activity was Tocused in the pH range 4-5 as shown by Ihe stippled area. Fractions in this region were pooled and rerocuscd to give a pH gradieni of 4-5 (lower trace). Fractions from this second experiment were dialy.sed extensively against distilled water, adjusted to a constant volume, and allergenic activity titrated in a cat sensitive subject.
Variations of the comparative immunoelectrophoresis technique of Wadsworth & Hanson (I960) were also used to demonstrate immunological identity between the principle protein in fractions 72 84 and cat serum albumin. In order to determine the mean molecular weights of the allergens, peak tubes from the two allergen rich regions (58 and 78) were labelled with '^'^1 by the chloraniine T method (Hunter & Greenwood. 1962) and fractionated on a calibrated column of Sephadex G-2(X) equilibrated in Tris buffer pH 7-6. Radioactivity was eluted in regions corresponding to molecular weights of 55,000 and 69,000 for the allergens of pl = 41 and pi = 435 respectively. In the case of allergen pi = 4 1 a minor peak of activity was also observed in fractions with an apparent molecular weight of 27,000. After dialysis against 0-1 M acetic acid.. 25 niM sodium chloride (pH 2-2) the '^''I-
Cat allergens
247
labelled allergen pi = 4-1 was fractionated on Sephadex G-200 equilibrated in the same buffer and no change in elution position was observed. However, when the labelled allergen (pl^4-l) was subjected to reduction in 0-1 M mercaptoethanol for 1 hr and then fractionated under identical conditions, the radioactivity eluted in the molecular weight region 6000-15,000 (see Fig. 6). The amino acid and carbohydrate compositions of allergen pi = 4-1 (fractions 57, 58 and 59, Fig. 2) are given in Tables ! and 2.
Fig. 3. Double diffusion analysis of Tractions obtained by isoelectric Tocusing of cat allergen extract (Fig. 2, lower trace): 48-120, fractions from the isoelectric profile; A, allergen active pool from Fig. I: B, allergen active poo! from Fig. 2 (upper trace): a-cc, rabbit anti-cat extract antiserum. Arrows indicate precipitin lines which correlate cJoscly with skin test activity.
There was insufficient allergen pi = 4 1 for use in the radioallergosorbent assay and a mixture of the two allergens (pi 41 and pi 4-35) obtained after Sephadex G-200 gel filtration (see Fig. I) was therefore used. In a study of adults classified as skin test negative or skin test positive (using the Bencard prick test solution) a significantly higher level of specific IgE binding was observed in the cat sensitive subjects (f
Allergen pl:4-i*, (minor peak) Mol. wt 27,000 Allergen p l ^ i - l * i imajor peoh) J Mol. wt 55,000
100
Allergen
Mol. wt 69,000
3-5
40
4-5 Log mol. wt
50
Fig. 6. Apparent molecular weights ofcat allergens obtained using Andrews (!964) plot of standardizing proteins. Assumed mol. wts for standards were as follows: IgG {150,000), human serum albumin (69,000) and pFc' fragment of human IgG (27,000). At neutral pH* cat allergen pi = 4-35 eluted with a mol. wt of 69,000 and cat allergen pi = 4-1 with a mol. wt of 55,000. A minor peak of allergen pl = 4-l also eluted with a mol. wt of 27,000. After reduction, alkylation and dissociation! the pl = 4-l allergen eluled within lhe mot. wl range 600015,000.
allergens as 24,000 70,000 (Hoiford-Strevens, 1973). 30.000-60.000 (Ohman et al., 1973) and 32.000 (Brandt ct ai, 1973) but these figures were obtained using unfractionated extracts which may have contained more than one allergen. The molecular weight ofcat allergen pl = 41 labelled with V^^ (55.000) was not changed after extensive dialysis in a dissociating bufier, pH 2 2. However, after reduction with mercaptoethanol and alkylation. all the radioactivity was in a peak eluted in the molecular weight region 6000-15.000. These observations are consistent with an allergen subunit structure of four polypeptide chains (mol. wt approx. 13,000) linked by disulphide bridges. According to this model the allergen exists primarily as a tetramer of molecular weight 55,000 and, possibly as a result of denaturation during isolation, can also occur as a dimer (mol. wt 27,000). Such an allergen would satisfy the prerequisite for activity suggested by Stanworth (1973), i.e. 'a dimeric structure is an essential requirement for allergenicity'.
Cat allergens
251
Table 1. Amino acid composition ofcat allergen pl=4-l Atnino acid*
Ragweed IgG iCyi Cat allergen pl = 4-l Cod allergen Wl (residues allercen E domain)ll per 109 (residues Number of Nearest residues)!: per 108 residues per four integer residues of residues) S phenylalanine
Lysine Histidine Arginine Aspartic acid Threoninet Serinef Glulamic acid Proiine Glycine Alanine Vallne Methionine Isoteucine Leucine Tyrosine Phenylalanine
6-8 2-2 4-2 98
10 2 100
15 0 50
12 0 10 4 5-6 2-4 3-4 62 1-8 40
7 2 4 10 10 10 15 5 12 10 6
2 3 6 2 4
12 0 1 13 2 6 10 0 10 25 6 0 5 8 I !0
6
2 5 16 6 9 8
5 12 10 2
7 7 I 4
9 3 3 10 8 14 14 9 6 3
8 2 2 9 5 4
• Trypiophan and half-cysline nol determined. t Corrected values. J Data of Elsayed ci al. (1972) modified for comparison. § Data of King, Norman & Tao (1974) modified for comparison. II Protein Eu sequence, residues 334-446 (Edelman el at, 1969).. Table 2. Carbohydrate composition of cat allergen pi = 4'I Sugar Mannose Galactose Glucose Glucosamine Fucose Sialic acid Glucosaminic acid
Content (n moles/mg allergen) 8-8 48 161 07
Trace Trace Trace
The amino acid composilion of cat allergen pi =4-1 was calculated assuming a total amino acid content of 108 residues (moi. wt 12,000, i.e. a subunit of the putative tetramer) and in Table 1 is compared with the amino acid composition of a single hom*ology region (CH3) of the human y-chain of IgG which has 109 residues. Such a comparison shows the allergen to be relatively deficient in proiine, a residue which is known to be involved in the coiling of peptide chains {Springall, 1954) and which
252
C. R. Stokes and M. iV. Turner
probably contributes to the compactness of the immunoglobulin domain. In contnsi the allergen is very rich in the neutral amino acids glycine and aiunine which occur less frequently in the Cy3 hom*ology region. The^ ratio of acidic to basic amino acids for lhe cat allergen is 1-9 and for the immunoglobuUn subunit, 1-7. Comparison of lhe amino acid composition of the cat allergen with two other well characterized allergens (Table I) shows that the codfish allergen M has no proiine and ragweed antigen E is also relatively deticicnl in this residue. Both allergens are also comparatively rich in glycine and alanine.
>
Skin tesT
Fig. 7. Radioallergosorbcnl test for IgE antibodies in adulls classified as skin lest negative or skin test positive using Hcncard cat prick test soliilion. Mixed cat allergens (pi =4-1 and pl = 4'35) ob* taincd after Sephadex G-200 gel filtration (Fig. 1) were coupled to Sigma cellulose particles.
The major sugars detected in the cat allergen pi = 4-1 were mannose, galactose and glucose. Both galactose and mannose have been detected in horse dandruff and galactose, mannose and glucose in human dandrulT(Berrens. 1971) but it is not known whether these sugar residues contribute to the allergenicity of the epithelial danders. It is likely that there are several distinct cat epithelial allergens. We have used a single cat sensitive subject for skin testing and the recognition of allergens is clearly restricted by the reactive capacity of the individual used. Nevertheless when the combined antigens (pl = 4-l and pI = 4-35) were used in a RAST procedure the serum of most cat sensitive individuals showed significant levels of binding. This suggests that both cat serum albumin and the immunochemically characterised allergen pi = 4 1 arc of major importance in cat sensitivity. Ackniiwlcdgnicnts We thank Miss L. Graham for her co-operation in the skin test assays and Dr H.
Cat allergens
253
Bennich, Institute for Molecular Biology. Aarhus University, Aarhus. Denmark for the amino acid and carbohydrate determinations. This work was supported by the Chest and Heart Association. References ANDREWS, P. (1964) Estimation of the molecular weights of proteins by Sephadex gel-fi!tration. Biochi'iiiical Journal, 91, 222. BF-LIN, L . (1972) Immunological analyses of birch pollen antigens, with special references toallergenic components, hilernatiotml Arvhires of Allergy and Applied ti?iinun(>!o!,'y, 42, 300. BERRENS, L . (1971) The Chemistry of Alopic Allergens, Mononriiphs in allergy, 7, Karger, Basel. BRANDT, R . , PONTERIUS, G . & YM.AN, L . (1973) The allergens oi cat epithetia and cat serum, tniernational Areliivcs of Allergy and Applied Ininuiroloffy, 45, 447. Cl-AMi', J.R., DAWSON. G . & HOUGH, L . (1967) The simultaneous estimation of 6-deoxy-L-galaclosc (L-fucose), D-mannose. D-galactose, 2-acetamiclo-2 deoxy-D-gluco.sc (n-acctyl-D-gliicosamine) and acetylneuraminic acid (Sialic acid) in glycopeptides and glycoproteins. Biochimka el blopliysica acla, 148, 342. co*kE, F . (1927) Review of 1000 cases of asthma. Brilish MedicaUournal, i, 955. EDSLMAN, G.M.,
CUNNINGHAM, B.A., G A L L , P.D., RUTISHALISKR, U . & WAXDAL, M.J. (1969) The
covalent structure of an entire / G immunoglobulin molecule. Proeecdings of lhe National Academy of Sciences of the U.S.A., 63, 78. ELSAYED, S., AAS, K . , SLKTTCN, K . & JOHANSSON, S . G . O . (1972) Tryptic cleavage of a hom*ogeneous
cod fish allergen and isolation of two active polypeptide fragmenls. Immunochemistry, 9, 647. FHANKI.ANI>. A . W . (1967) Allergy to animals. Acta allergolica. 22, 175. GRABAK, P . & WILLIAMS, C . A . (1953) Methode permettant Tetude conjugee des proprieles electrophoretiques et Immunochimiques d'un melange de proteines. Application au s6rum sanguin. Biochimica et biophysica acta, 10, 193. HOLFORD-STRFVENS, V. (1973) Allergenic activity ofcat and dog skin fractions obtained by Sephadex gel filtration. Clinical Allergy, 3, 225. HUNTER, W.M. & GREENWOOD. F . C . (1962) Preparation of Iodine-i3l labelled human growth hormone of high specific activity. Nature, 194, 495. KING, T. P., NORMAN, P.S. & TAO, N . (1974) Chemical modifications of the major allergen of ragweed pollen, antigen E. tmmiinoehcnustry, I I , 83. OHMAN, J.L., LOWELL, F.C. & BLOCH, K.J.( 1973) Allergens of mammalian origin: Characterisation of allergen extracted from cat pelts. Journal of Allergy and Clinical hnmunology, 52, 231. OHMAN, J.L., LOWELL, F.C. & BLorH,K.J. (1974) Allergens of mammalian origin. III. Properties of a major feline allergen. Journal of immunology, 113, 1668. OUCHTERLONY, O.( 1958) Difiusion-in-gel methods for immunological analysis./*rOjprf5.j(n./4//ef^^,5,l. PONTERIUS, G., BRANDT, R . , HULTEN, E . & YMAN, L . (1973) Comparative studies on iho allergens of
horse dandrutT and horse serum, inlernaliomd Archives of Allergy and Applicil Immunology. 44, 679. SALTER, H . H . (1864) On Asthma its Pathology and Tieatmi'iil. J. and A. Churchill, Ltd, London. SPACKMAN, D.H., STEIN, W . H . & MOORE, S. (1958) Aulomatic recording apparatus for use in lhe chromatograpiiy of amino-acids. Analylical Cliemisfry, 30, 1190. SPRINGALL, H . D . (1954) The Structural Chemistry of Protein.',: Bulterworlhs, London. STANWORTH, D . R . (1957a) The isolation and identification of horse-dandrulf allergen. Bioehcmical Journal, 65, 582. STANWORTH, O . R . (1957b) The use of lhe gel-precipitation technique in the idcntilicalion of horse dandrulT allergen, and in the study of the serological relationship between horse dandruff and horse serum proteins. International Archives of Allergy and Applied Immunology, U , 170. STANWORTH, D.R. (1973) Immediale hypersensilivity. Frontiers of Biology, vol. 28. Norlh-Holland. Amsterdam. TURNER, M . W . & BENNK H, H . (l968)Subfragments from the Fc fragment of human immunoglobulin G. Isolalion and physicochemical characterisation. Biochemical Journal, 107, 171. VAN HoFTSEN, B., VAN K.LEY, H. & FAKER, D . (1965) An extracellular proteolytic enzyme from a strain of Arthrobaeter. 11. Purification and chemical properties of the enzyme. Biothimica el biophysiea acta, 110, 585.
254
C. R. Stokes and M. W. Turner
VARGA, J . M . & CESKA. M . (1972) Characterisation of allergen extracts by polyacrylamide gel isoeiectricfocusing and radioimmunosorbent allergen assay. II. Dog and cat allergens. Inter-
national Archives of Allergy and Applied immunology, 42, 4.18. WADSWORTH, C. & HANSON, L.A. (1960) Comparative analysis of immune electrophoretic precipitates employing a modified immune electrophoretic technique, international Archives of Allergy and Applied immtinoiogy, 17, 165, WALKER, I.C. (1917) Studies on the sensitization of patients with bronchia! asthma to the different proteins found in the dandruff of the horse and in the hair of the cat and ihe dog and to Ihe sera of these animals. Journal of Medical Research, 35, 497. E, L., BFNNirn, H. & JOHANSSON, S . C O . (1967) Diagnosis of allergy by an in vitro test for allergen antibodies. Lancet, ii, 1105. R.P, (1917) Study (VI). Immunochemical studies of the proteins ofcat hair. Jotumd of immunology, 2,227. YMAN, L., BRANDT, R. & PONTLRIUS, G . (1973) Serum albumin—an important allergen in dog
epithelia extracts, tnternational Archive.'! of Allergy and Applied immunology, 44, 358.
