Journal of lrnmunological Methods, 156 (1992) 61-67 © 1992 Elsevier Science Publishers B.V. All rights reserved 0022-1759/92/$05.00
Preparation of monoclonal mouse antibodies against two specific eu-melanin related compounds A. K a m m e y e r a, L.A. O o m e n
and S. Pavel b
a Department of Dermatology and Clinical Chemistry, Unit'ersity of Amsterdam, Academisch Medisch Centrum, Amsterdam, Netherlands, and b Department of Dermatology, Unit'ersity Hospital of Leiden, Leiden, Netherlands (Received 26 March 1992, revised received 2 June 1992, accepted 15 June 1992)
Two eu-melanin precursors, 6-hydroxy-5-methoxyindole-2-carboxylicacid (HMI2C) and 5,6-dihydroxyindole-2-carboxylic acid (DHI2C) were synthesized and coupled to bovine serum albumin, hemocyanin and polylysine by the combined action of carbodiimide and succinimide. These indole-carrier conjugates served as antigens for the production of specific antibodies against DHI2C and HMI2C in B A L B / c mice. The specificity of these antibodies was tested using a combination of affinity chromatography and ELISA procedures. Polyclonal mouse antibodies reacted with the indole-carrier conjugates, but not with the unbound indole compounds. Monoclonal antibodies from two hybridoma cell lines were obtained from a HMI2C-immunized mouse after a fusion with four subclonings. They reacted with free HMI2C and to a lesser extent with unbound DHI2C. One monoclonal showed 50% inhibition in the ELISA test at concentrations of 0.6 /xmol.l -~ and 5 /~mol. I -~ for HMI2C and DHI2C, respectively. These antibodies did not show any cross-reactivity with nine structurally related compounds and should be valuable reagents for the detection and quantification of HMI2C and other eu-melanin related compounds. Key words: Monoclonal antibody; Eu-melanin; lndole compound
Introduction Correspondence to: A. Kammeyer, Neurozintuigen Laboratorium, K.2-Noord Academisch Medisch Centrum, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands. Abbre~'iations: BSA, bovine serum albumin; CAMOR, coupling agent modified residue(s); DHBSA, DHI2C coupled to BSA; DHHC, DHI2C coupled to hemocyanin; DHI2C, 5,6-dihydroxyindole-2-carboxylic acid; DHPL, DHI2C coupled to polylysine; EDC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide; ELISA, enzyme-linked immunosorbent assay; HMBSA, HMI2C coupled to BSA; HMHC, HMI2C coupled to hemocyanin; HMI2C, 6-hydroxy-5-methoxyindole-2-carboxylic acid; HMPL, HMI2C coupled to polylysine; PBS, phosphate-buffered saline pH 7.4; Sulpho-NHS, N-hydroxysulphosuccinimide (sodium salt).
Clinical oncology is in great need of specific and sensitive laboratory tests for diagnostic purposes and the follow-up of cancer patients. However, only a few tumours produce specific metabolites which can serve as tumour markers. In melanocytes and melanoma cells, synthesis of the polymeric pigment melanin is an integral part of the intracellular metabolism, that has attracted the attention of many clinical chemists and oncologists interested in tumour markers. Melanin precursors are produced in cellular compartments containing the enzyme tyrosinase, which converts
tyrosine into the eu-melanin precursors 5,6-dihydroxyindole (DHI) and 5,6-dihydroxyindole-2carboxylic acid (DHI2C). However, not all precursor molecules are involved in the polymerization process. Some of them escape from the melanogenic compartments (Pavel et al., 1983) and are partly methylated by catechol-O-methyltransferase (Smit et al., 1990). The methylated precursors, together with the 5,6-dihydroxyindoles leave the cells, to be excreted from the organism. Urinary excretion of these metabolites has been thoroughly studied (Pavel et al., 1981, 1984a, b). Experience has shown that only certain patients with malignant melanoma excrete increased quantities of the melanin-related substances in their urine (Pavel and Van der Slik, 1986). The determination of the urinary excretion of these tumour markers is therefore of limited value. In our opinion, their blood concentration might better reflect the extension of tumour load in the organism. Elevated blood levels of some indole metabolites in melanoma bearing mice can be detected by high performance liquid chromatography (Wakamatsu et al., 1990), although this technique lacks the sensitivity to detect all the indole metabolites in normal blood samples. An alternative method of detection is by an immunoassay, which does not require expensive
equipment and is easier to perform. In the present study a first goal has been achieved by the production of specific (monoclonal) antibodies against the eu-melanin precursors DHI2C and HMI2C. These antibodies may facilitate the development of an immunoassay which could be useful in studies of pigment cell abnormalities. Materials and methods
Materials Peroxidase-labeled goat anti-mouse immunoglobulin was obtained from Dako, Glostrup, Denmark. BSA, keyhole limpet hemocyanin and polyL-lysine were purchased from Sigma Chemical Co., St. Louis, MO, USA. EDC-hydrochloride and sulpho-NHS were obtained from Pierce Europe, Oud Beijerland, Netherlands. Interleukin-6 was purchased from the Central Laboratory of Blood Transfusion Services, Amsterdam, Netherlands. DHI2C was synthesized from 3,4-dihydroxyphenylalanine by mild oxidation and ethyl acetate extraction (Wakamatsu and Ito, 1988). Fine, grey crystals were obtained, with a melting point of 235°C. Molecular structure was confirmed by mass spectrometry (5,6-DHI2C(PFP)2-HFIP derivative, M+ = 635). 6-benzyloxy-5-methoxy-indole-2-carboxylic acid was obtained from Aldrich Benelux.
Hapten solutions 1% D H I 2 C in dioxane 1% HMI2C
dioxane Carrier compounds 1.3% BSA
_ 50 a
1 M EDC-HCI
0.5 M s u l p h o - N H S
Calculated from the averaged mass of amino acid residues.
63 This compound was catalytically hydrogenated (Benigni and Minnis, 1965) to yield HMI2C, m.p. 256°C (dec.) whose molecular structure was confirmed by mass spectrometry (6H5MI2C-PFPH F I P derivative, M ÷= 503). Fluorescence data were obtained using a Perkin Elmer 650-10S fluorescence spectrophotometer using wavelengths set for excitation at 295 nm and for emission at 385 nm. The slit width was 5 nm.
Conjugate preparation The micromolar quantities of all the reactants used in the preparation of conjugates are summarized in Table I. In addition to coupling agent EDC, another agent, sulpho-NHS, was used in order to enhance hapten labeling (Staros et al., 1986). Both coupling agents were added in excess in order to obtain sufficient yield. The reaction was carried out at room temperature in the dark for 6 h using solutions bubbled with nitrogen. The total reaction volume was adjusted to 5 ml with 0.05 M sodium phosphate buffer pH = 7.2. BSA, hemocyanin and polylysine were dissolved at concentrations of 0.1 M 'amino acid residue'. This expression was used, because the total molecular mass of hemocyanin and polylysine is rather undefined. The average molecular mass of the amino acid residues present in BSA and hemocyanin was taken as 64; that of polylysine was taken as 130. The reaction rate was monitored by HPLC. After 6 h the DHI2C and HMI2C peaks disappeared. Then, the reaction mixtures were dialysed against 2 x 1 liter phosphate-buffered saline and 3 x 1 liter purified water. Occasionally formed precipitates in the dialysis bags were removed by centrifugation. L a b e l ratios (Table III) were inferred from fluorescence measurements and from use of the biuret protein assay method. In contrast with some other protein assays, the polylysine concentration was readily measured using that method. The processed reaction mixture with BSA and HMI2C, but without E D C and sulpho-NHS, served as a control. Carriers modified by carbodiimide and succinimide, were also prepared for control use in the ELISA tests and as references in the fluorescence analysis of conjugate label. They were processed in a similar manner to that described for the coupling reaction, except that
HMI2C was omitted from the reaction mixture. The conjugates were frozen at - 2 0 ° C until further use.
Immunization Male B A L B / c mice were immunized with each indole-carrier conjugate by intraperitoneal injections (0.1 ml volume) of a 1:1 emulsion of Freund's complete adjuvant and the immunogen dissolved in water. A 1:1 mixture, in which the immunogen solution was replaced by saline, was administered to two control mice. Three injections using incomplete Freund's adjuvant were given after 1 month with weekly intervals. After the third immunisation injection the mouse antibody titre was determined by ELISA. Mouse blood was obtained by orbital puncture and after clotting and centrifugation serum was obtained. An immunization was considered to be successful when a read-out of 1.0 A U was obtained in the ELISA test with mouse antiserum diluted at least 4 × 104. If this was not obtained, one extra intraperitoneal injection followed. The mice with a sufficiently elevated antibody titre were killed and the spleens were freshly processed for fusion with the NS.1 myeloma cell line. Blood was collected, centrifuged and serum samples frozen in diluted form at -20°C.
Production of monoclonal antibodies Spleen cells were homogenized in a Potter tube and washed with fusion medium. Hybridoma cells were obtained after a fusion between spleen cells and myeloma cells at a ratio 5 : 1 in the presence of ultrapure polyethyleneglycol 1500. These cells were allowed to grow with recombinant IL-6 stimulant (100 U / m l ) in microtiter wells at 37°C in a 5% carbon dioxide atmosphere. The mixed cells from the fusion were seeded at a density of 105 spleen cells per well. Hybridoma cell growth was favored by the presence of hypoxanthine, aminopterin and thymidine (Sigma Chem. Co.) in RPMI 1640 growth medium supplied with 12% foetal calf serum (Gibco BRL, Breda, Netherlands). On day 4 the same amount (100/~1) of H A T medium was added to the wells as on the fusion day. On days 7 and 10 H A T medium was restored. Hybridoma media were screened using the ELISA proce-
64 dure. Positive wells were subcloned in order to obtain monoclonal hybridoma cell lines. Purification of monoclonal antibodies was performed by 50% ammonium sulphate precipitation, followed by a passage through Sepharose protein A (Pharmacia, Uppsala, Sweden).
ELISA test for specific antibodies All solutions were bubbled with nitrogen, except for washing fluid. Incubations were performed at 37°C in the dark, each followed by a wash of 0.05% Tween 80 in pure water. Microtitre plates (96 wells) were coated with 40 /zg. 1-1 HMPL in 0.02 M carbonate buffer pH 9.6 for 90 min. Modified polylysine (see conjugate preparation) and normal mouse serum served as controls. Residual binding sites were blocked by incubation with 2% BSA in PBS for 60 min. The mouse antibody, or the hybridoma medium dilutions in PBS, containing 2% BSA and 0.05% Tween 80, were allowed to react for 90 min. Prior to this stage, any potentially competitive test solution was added to the well in order to test for inhibition. The total volume of the well was kept constant (100/zl). Horseradish peroxidase labeled goat anti-mouse immunoglobulin (Dako, Glostrup, Denmark) was used at a 1000-fold dilution in PBS with 2% BSA and 0.05% Tween 80 over an incubation period of 60 min. Tetramethylbenzidine and hydrogen peroxide were then added and after 10 min colour development was stopped by the addition of 2 M sulphuric acid. Colour intensity at 450 nm was determined using an EIA reader (Bio-Rad, Richmond, CA, USA).
Subtyping of monoclonal anti-HMI2C antibodies The subclass and light chain type of the monoclonal antibodies were determined from the hybridoma medium using a test kit from Boehringer Mannheim, Germany.
Thin layer chromatography Silica plates (layer thickness 0.25 mm) with fluorescent indicator were used to evaluate the covalent linkage of the haptens. 1 - 4 / x l samples were eluted with an acetone, ethyl acetate, water, acetic acid mixture used in a 8 : 8 : 3 : 1 ratio, respectively.
Specificity tests with a combined affinity chromatography-ELISA experiment All solutions were bubbled with nitrogen and all manipulations were performed in test tubes, except for the recording of colour intensities.
Preparation of Sepharose-spacer arm-ligand. 1,2-diaminoethane was coupled to CNBr-Sepharose according the manufacturer's instructions (Pharmacia, Uppsala, Sweden) to yield a Sepharose-spacer arm complex. After three washing cycles with pure water, the ligand (HMI2C) was linked to the complex by EDC at neutral pH. The mixture was allowed to react overnight at room temperature in the dark. The Sepharose-spacer arm-ligand complex (antigen complex) was thoroughly washed with the following sequence of solutions: 10% dioxane, water, 0.5 M NaC1, water and PBS with 0.05% Tween 80. ELISA procedure. Residual binding sites of the antigen complex were blocked by incubation at room temperature for 60 min with PBS, containing 2% BSA and 0.05% Tween 80. The beads were washed with PBS, containing 0.05% Tween 80. Then, dilutions of anti-HMHC, anti-HMPL and normal mouse serum in PBS with BSA and Tween 80 were added to the antigen complex and incubated for 120 min. The serum dilutions are given in Table IV. The antigen complex was then again washed with PBS, containing 0.05% Tween 80 before the addition of diluted (1/1000) goat anti-mouse immunoglobulin with peroxidase label. After an incubation period of 120 min, the antigen complex was washed three times with PBS containing 0.05% Tween 80. Tetramethylbenzidine and hydrogen peroxide were added and colour development stopped after 2 min by the addition of 2 M sulphuric acid. The samples were centrifuged immediately and the supernatants quickly transferred to ELISA plate wells. Yellow colour intensities were recorded as usual on an EIA reader. Normal mouse serum and Sepharose-spacer arm complex without indolyl ligand served as controls. Results
A fusion yielded 77 hybridoma cell lines, which secreted antibodies directed against the hapten
T A B L E II
CROSS-REACTIVITY A N T I - H M I 2 C MAb.
0, - J J
nmol HMI2C or DHI2C per litre
Fig. 1. E L I S A results for competitive inhibition of i m m u n e complex formation between coated H M P L (40 / x g . 1 - l ) and monoclonal a n t i - H M H C antibodies by dilutions of HMI2C (11 II) and D H I 2 C (A A). The control was pure water.
site of the HMPL conjugate. Two cell lines of this fusion secreted antibodies, which also reacted with HMI2C in the free form. The purified antiHMI2C monoclonal antibodies (20 D 12 and 11 H 11) only differed in their detection limits for the free hapten in our ELISA. Fig. 1 shows the inhibitory effect of HMI2C and DHI2C on immune complex formation with monoclonal antiHMI2C antibody (20 D 12). 50% inhibition of immune complex formation was achieved with 0.6 /zmol • 1-1 HMI2C and 5/zmol • 1 - 1 DHI2C. The standard deviation of an ELISA without 'free' HMI2C was 0.033 absorbance units. Significant inhibition (2 × SD) occurred at 0.05 /xmol.1-1 and at 0.7 /zmol. 1-1. Table II illustrates the absence of any cross-reactivity for nine structurally related naturally occurring compounds. Only the target compounds HMI2C and DHI2C showed substantial inhibition. Table III shows the hapten labeling of the conjugates, obtained by the combined use of carbodiimide and the succinimide. The conjugation of DHI2C with carrier compound yielded lower hapten labeling than was the case with HMI2C. The highest degree of bound fluorescent hapten was obtained with polylysine as a carrier. Without the coupling agents the labeling was negligible
C o m p o u n d tested for crossreactivity (final concentration 1 l / z m o l . 1-1)
Water 5-hydroxyindole-3-acetic acid 5-hydroxytryptamine 3,4-dihydroxyphenylalanine 3,4-dihydroxyphenylet hylamine Norepinephrine Normetanephrine Vanillylmandelic acid Tryptophan Guanine 5,6-dihydroxyindole-2-carboxylic acid (DHI2C) 6-hydroxy-5-met hoxyindole-2-carboxylic acid (HMI2C)
0 - 13 - 6 - 5 - 6 - 9 - 10 - 8 - 8 - 17
+ 41 + 82
(control). Alternative coupling conditions, such as the Mannich formaldehyde reaction, bifunctional linking with N-(4-aminobenzoyl)-N'-(pyridyl-dithiopropionyl)hydrazine or acetylation of indolic compounds prior to EDC and sulpho-NHS coupling action, only yielded one hapten molecule to more than 4 × 103 amino acid residues. Mouse anti-HMHC antiserum dilutions of 38 × 103 and 120 × 103 resulted in absorbance values of 1.0 in ELISA against coated HMPL at 4 p,g. 1-1 and 40 /.tg. 1-1, respectively. Unreliable results were obtained at a coating concentration of 0.4/zg HMPL" 1- t. Among the possible combi-
T A B L E III THE EXTENT OF DHI2C AND HMI2C BINDING TO CARRIER COMPOUND Conjugate code
Bound D H I 2 C or HM12C (/~mol.1-1 )
Averaged amino acid residue (mmol • 1- I )
N u m b e r of bound indole molecules to 104 amino acid residues
DHBSA DHHC DHPL HMBSA HMHC HMPL Control
6.4 1.8 1.3 16 5.3 11 0.1
l0 7.0 2.0 11 2.9 3.1 3.6
6.4 2.6 6.5 14.5 21.3 35.7 0.2
66 TABLE IV SEPHAROSE-ELISA EXPERIMENT Mouse sera
Readings-blank Sepharose-spacer arm complex + Ligand
1000 4000 16000
0.59 0.68 0.44
0.15 0.18 0.19
8000 32000 128 000
1.11 0.59 0.20
0.44 0.15 0.06
1000 4000 16000
0.25 0.10 0.06
0.23 0.09 0.06
Key: Sample data are shown in bold; control data are shown
TLC experiments revealed the formation of a fluorescent reaction product ( R f = 0.29) and a weaker spot (Rf = 0.39) from the action of EDC on HMI2C. On the other hand, TLC experiments with sulpho-NHS and HMI2C showed partial conversion of the indole compound into a nonfluorescent moiety ( R f = 0 . 0 6 ) and unreacted HMI2C (Rf = 0.89) could still be detected. A spot (Re= 0.61) developed following the mixture of HMI2C with both coupling agents. No HMI2C could be detected and unelutable material was formed. When the reaction time exceeded 5 h several other reaction products were detected.
in normal typeface.
nations, the H M P L / a n t i - H M H C complex resulted in the highest absorbane in the ELISA procedure. HMI2C and DHI2C did not significantly interfere with immune complex formation using a polyclonal antibody source. However, a moderate inhibition of H M P L / a n t i - H M H C complex formation was observed when HMI2C, preincubated with EDC and sulpho-NHS, was added to the ELISA wells. The lowest affinity was observed with coated HMBSA (400/xg. l - l ) reacted against anti-HMPL. In this case an absorbance level of 1.0 was obtained at an antiserum dilution of 4 x 103. Polylysine modified coupling agent, used at a concentration of 40 p.g.1-1, showed no interference with antiHMHC. The formation of immune complexes using mouse anti-hapten antibodies and HMI2C as a ligand on Sepharose beads (sample) is illustrated in Table IV. Controls without the indolyl ligand and ligand beads with normal mouse serum, showed low peroxidase activities. These results suggested that the antibodies were mainly directed against the hapten itself. Additional results Subtyping the monoclonal antibodies using an ELISA procedure demonstrated that the antibody belonged to the IgG1 subclass and contained K light chains.
Research over the past decade has shown that indolic eu-melanin precursors can be found at elevated levels in the urine and blood of melanoma patients (Pavel and Van der Slik, 1986; Wakamatsu et al., 1990). These substances comprise: 5,6-dihydroxyindole, 5-hydroxy-6-methoxyindole, 6-hydroxy-5-methoxyindole (non-carboxylated indoles) and 5,6-dihydroxyindole-2-carboxylic acid (DHI2C), 5-hydroxy-6-methoxyindole-2-carboxylic acid and 6-hydroxy-5-methoxyindole-2-carboxylic acid (HMI2C). Our choice of the carboxylated indoles for the production of antibodies was based on their known chemical properties. The 2-carboxyindoles are less susceptible to oxidation and polymerization than the non-carboxylated indolic compounds. Furthermore, the 2-carboxyindoles exhibit an intense fluorescence, which makes them suitable for specific detection. The carboxylic acid group at the 2 position facilitates the formation of a molecular bridge with a free amino group of the carrier compound. For the coupling reaction, polylysine was selected because of its high level of potential coupling sites, namely the amino groups of the lysine monomers. The conjugate of polylysine with HMI2C showed the highest degree of labeling (HMPL, Table III). Another selected carrier was hemocyanin which had the potential to elicit a strong antibody response. Indeed the HMHC conjugate showed the highest anti-hapten response and the
anti-HMHC antiserum had to be diluted 120 x 103 for a proper read-out in the ELISA. The use of BSA as a carrier resulted in intermediate results with respect to both labeling and antibody response. All the conjugates were prepared using carbodiimide, and succinimide agents after several attempts with other coupling conditions. However, some disadvantages were noted when synthesizing a covalently linked hapten. TLC experiments showed the conversion of HMI2C into unknown substances in the presence of the coupling agents. The bright fluorescence "spot of HMI2C disappeared almost completely after a reaction period of 2 h and one of the TLC spots did not exhibit any fluorescence. The corresponding compound could just as well be linked to carrier compound in order to produce a non-fluorescent label. Moreover, the fluorescence of other labels may be quenched by surrounding polymer chains of the carrier. These two phenomena may explain the poor labeling (Table III), calculated from the fluorescence data. The presence of carrier compound together with HMI2C and coupling agents resulted in virtually the same TLC pattern. It is possible that a significant part of the hapten label of the conjugates had an altered structure, as was also inferred from the immunologic properties of the mouse antibodies raised. After immunization the ELISA test revealed high (polyclonal) antibody titers against the hapten. However, the addition of HMI2C in the free form did not inhibit a competitive ELISA whereas the addition of HMI2C, preincubated with the coupling agents, did show some inhibition (10-15%) in a competitive ELISA. In the course of monoclonal antibody production a few hybridoma clones did secrete anti-HMI2C antibodies. We assume therefore that a part of hapten label retained its original structure. The Sepharose-ELISA experiment represented a control, confirming the lack of antibody reactivity towards a molecular bridging group. Further controls excluded possible interference from CAMOR compounds (Briand et al., 1985) in the formation of immune complexes. Carrier compounds treated with the coupling agent did not react with either the poly- and monoclonal antibodies. In the next phase of this research we shall address the detection of HMI2C
and DHI2C in the blood of healthy individuals and of patients with malignant melanoma.
Acknowledgement This research was partly supported by the Dutch Cancer Society. (project IKA 90-20). We are grateful to dr. T.A. Eggelte of the Royal Tropical Institute for his valuable advice.
References Benigni, J.D. and Minnis, R.L. (1965) The synthesis of 5,6-dihydroxyindole and some of its derivatives (1). J. Heterocycl. Chem. 2, 387-392. Briand, J.P., Muller, S. and van Regenmortel, M. (1985) Synthetic peptides as antigens: Pitfalls of conjugation methods. J. Immunol. Methods 78, 59. Pavel, S., Muskiet, F.A.J., Nagel, G.T., Schwippelova, Z. and Duchon, J. (1981) Identification of two Thormahlen-positive compounds from melanotic urine by gas chromatography-mass spectrometry, J. Chromatogr. 222, 329-336. Pavel, S., Muskiet, F.A.J., De Ley, L., The, T.H. and Van der Slik, W. (1983) Identification of three indolic compounds in a pigmented melanoma cell culture supernatant by gas chromatography - mass spectrometry, J. Cancer Res. Clin. Oncol. 105, 275-279. Pavel, S., Boverhof, R. and Van der Slik, W. (1984a) Identification of Thormahlen positive compound 'B' in urine of patients with malignant melanoma. Arch. Dermatol. Res. 276, 156-159. Pavel, S., Boverhof, R. and Wolthers, B.G., (1984b) Identification of 5-hydroxy-6-indolyl-O-sulfate in urine of patients with malignant melanoma. J. Invest. Dermatol. 82, 577579. Pavel, S. and Van der Silk, W. (1986) Analysis of eu-melanin related compounds in urine by HPLC with fluorimetric detection, J. Chromatogr. 375, 392. Smit, N., Pavel, S., Kammeyer, A. and Westerhof, W. (1990) Determination of catechol-O-methyltransferase activity in relation to melanin metabolism using high-performance liquid chromatography with fluorimetric detection. Anal. Biochem. 190, 286-291. Staros, J.V., Wright, R.W. and Swingle, D.M. (1986) Enhancement by N-hydroxysulfosuccinimide of water-soluble carbodiimide-mediated coupling reactions. Anal. Biochem. 156, 220. Wakamatsu, K. and Ito, S. (1988) Preparation of eu-melanin metabolites 5,6-dihydroxyindole, 5,6-dihydroxyindole-2carboxylic acid and their O-methyl derivatives. Anal. Biochem. 170, 335. Wakamatsu, K., Ito, S. and Fujita, K. (1990) Production, circulation and excretion of eu-melanin related metabolites in B-16 melanoma bearing mice. Acta Dermatol. Venereol. 70, 367.