Mercuric chloride-induced spin or ligation state changes in ferric or ferrous human cystathionine beta-synthase inhibit enzyme activity

Cystathionine beta-synthase is a key heme and pyridoxal phosphate-dependent enzyme involved in homocysteine metabolism in humans. The role of the recently discovered heme in this protein remains an important open question. The axial ligands to the heme in both the ferrous and ferric states have been assigned as cysteine and histidine residues, respectively. In this study, we have examined the effect of ligation and spin state changes in the heme on the activity of the enzyme. Treatment of the ferric enzyme with HgCl2 results in the conversion of six-coordinate low-spin heme to five-coordinate high-spin heme and is paralleled by a loss of activity. In contrast, treatment of the ferrous enzyme with HgCl2 results in replacement of the cysteine ligand by an unidentified sixth ligand and retention of the six-coordinate state, and is also accompanied by loss of enzyme activity. Treatment of the five-coordinate HgCl2-treated enzyme with thiols, such as homocysteine, results in reversion to a six-coordinate state. Resonance Raman spectroscopy with 34S-labeled enzyme reveals the return of the endogenous thiol ligand under these conditions and rules out direct coordination by the thiolate of homocysteine to the heme.

Characterization of NO binding to human cystathionine beta-synthase: possible implications of the effects of CO and NO binding to the human enzyme

Homocysteine is a key junction metabolite that can be converted to cystathionine in a reaction catalyzed by the heme and pyridoxal phosphate-dependent cystathionine beta-synthase. The heme has unusual spectroscopic properties and the axial ligands have been assigned as histidine and cysteine, respectively. Its role in the protein is not obvious from the chemistry of the beta-replacement reaction that is catalyzed. We have characterized the binding of the gaseous signaling molecule, NO, to cystathionine beta-synthase and examined its effect on the reactions catalyzed by the truncated dimeric form of the enzyme, W409X, which is a natural variant. Binding of NO appears to result in the formation of a five-coordinate ferrous nitrosyl species in which both endogenous ligands have been lost. This is in contrast to CO binding which is reported to displace the thiolate ligand and form a six-coordinate species. NO binds to the full-length enzyme with a K(d) of 281+/-50 microM and to the truncated enzyme with a K(d) of 350+/-44 microM. Binding of NO to the full-length enzyme inhibits activity with a K(i) of 320+/-60 microM. These studies demonstrate that as with CO, perturbation of the heme environment by NO is communicated to the active site with concomitant inhibition of enzyme activity, and suggests a regulatory role for heme in cystathionine beta-synthase.

Resonance Raman characterization of the heme cofactor in cystathionine beta-synthase. Identification of the Fe-S(Cys) vibration in the six-coordinate low-spin heme

Human cystathionine beta-synthase (CBS) is an essential enzyme for the removal of the toxic metabolite homocysteine. Heme and pyridoxal phosphate (PLP) cofactors are necessary to catalyze the condensation of homocysteine and serine to generate cystathionine. While the role for the PLP cofactor is thought to be similar to that in other PLP-dependent enzymes that catalyze beta-replacement reactions, the exact role for the heme remains unclear. In this study, we have characterized the heme cofactor of CBS in both the ferric and ferrous states using resonance Raman spectroscopy. Positive identification of a cysteine ligand was achieved by global (34)S isotopic substitution which allowed us to assign the nu(Fe-S) for the six-coordinate low-spin ferric heme at 312 cm(-1). In addition, the CO adduct of ferrous CBS has vibrational frequencies characteristic of a histidine-heme-CO complex in a hydrophobic environment, and indicates that the Fe-S(Cys) bond is labile. We have also found that addition of HgCl(2) to the ferric heme causes conversion of the low-spin heme to a five-coordinate high-spin heme with loss of the cysteine ligand. The present spectroscopic studies do not support a reaction mechanism in which homocysteine binds directly to the heme via displacement of the Cys ligand in the binary enzyme complex, as had been previously proposed.

Magnetic field effects on coenzyme B12-dependent enzymes: validation of ethanolamine ammonia lyase results and extension to human methylmalonyl CoA mutase

Enzymes with radical-pair intermediates have been considered as a likely target for purported magnetic field effects in humans. The bacterial enzyme ethanolamine ammonia lyase and the human enzyme methylmalonyl-CoA mutase catalyze coenzyme B12-dependent rearrangement reactions. A common step in the mechanism of these two enzymes is postulated to be homolysis of the cobalt-carbon bond of the cofactor to generate a spin-correlated radical pair consisting of the 5'-deoxyadenosyl radical and cob(II)alamin [Ado. Cbl(II)]. Thus, the reactions catalyzed by these enzymes are expected to be sensitive to an applied magnetic field according to the same principles that control radical pair chemical reactions. The magnetic field effect on ethanolamine ammonia lyase reported previously has been corroborated independently in one of the authors' laboratory. However, neither the human nor the bacterial mutase from Propionibacterium shermanii exhibits a magnetic field effect that could be greater than about 15%, considering the error limit imposed by the uncertainty of the coupled assay. Our studies suggest that putative magnetic field effects on physiological processes are not likely to be mediated by methylmalonyl-CoA mutase.

Coronary artery bypass grafting performed with or without a bypass pump: early results

INTRODUCTION

Traditionally, heart bypass surgery has required stopping of the heart and the use of cardiopulmonary bypass. Numerous complications have been associated with exposure to this extracorporeal circuit. Newer techniques of local cardiac wall stabilization now enable this operation to be performed safely "Off Pump". The early clinical results of Off Pump Coronary Artery Bypass (OPCAB) will be compared to a similar group of traditional Coronary Artery Bypass Grafting (CABG) patients.

METHODS

A retrospective review of 137 consecutive patients undergoing elective coronary artery bypass grafting was performed, 68 of who underwent traditional CABG and 69 of who underwent OPCAB. Inclusion criteria consisted of first time cardiac surgical procedures with an ejection fraction > or = 20%, without significant renal failure (creatinine < 2.0).

RESULTS

There was no statistical difference in the age, sex, cardiac function or underlying co-morbidities between those undergoing CABG and OPCAB. CABG patients had slightly more vessels bypassed than those in the OPCAB group (3.0 vs 2.6, p = 0.010). Despite similar preoperative characteristics, the OPCAB group experienced a reduction in morbidity without an increase in mortality.

CONCLUSION

In similar patient populations, OPCAB was associated with significantly reduced transfusion requirements, intubation time, ICU and overall hospital lengths of stay, with no increase in mortality. Further investigation is warranted to ascertain the role of the OPCAB in the general cardiac surgical community.

Assignment of enzymatic functions to specific regions of the PLP-dependent heme protein cystathionine beta-synthase

Cystathionine beta-synthase is a unique heme protein that catalyzes a pyridoxal phosphate (or PLP)-dependent beta-replacement reaction. The reaction involves the condensation of serine and homocysteine and constitutes one of the two major avenues for detoxification of homocysteine in mammals. The enzyme is allosterically regulated by S-adenosylmethionine (AdoMet). In this study, we have characterized the kinetic, spectroscopic, and ligand binding properties of a truncated catalytic core of cystathionine beta-synthase extending from residues 1 through 408 in which the C-terminal 143 residues have been deleted. This is similar to a natural variant of the protein that has been described in a homocystinuric patient in which the predicted peptide is 419 amino acids in length. Truncation leads to the formation of a dimeric enzyme in contrast to the tetrameric organization of the native enzyme. Some of the kinetic properties of the truncated enzyme are different from the full-length form, most notably, significantly higher K(m)s for the two substrates, and loss of activation by AdoMet. This is paralleled by the absence of AdoMet binding to the truncated form, whereas four AdoMet molecules bind cooperatively to the full-length tetrameric enzyme with a K(d) of 7. 4 microM. Steady-state kinetic analysis indicates that the order of substrate addition is important. Thus, preincubation of the enzyme with homocysteine leads to a 2-fold increase in V(max) relative to preincubation of the enzyme with serine. Since the intracellular concentration of serine is significantly greater than that of homocysteine, the physiological significance of this phenomenon needs to be considered. Based on ligand binding studies and homology searches with protein sequences in the database, we assign residues 68-209 as being important for PLP binding, residues 241-341 for heme binding, and residues 421-469 for AdoMet binding.

Vital immunohistochemical staining for a novel method of diagnosing micro-cancer. Examination of immunohistochemical staining of non-fixed fresh tissue

It becomes possible to establish a novel diagnostic method for micro-cancer by modulating the signals from the lesion, if lesions can be labeled with substances which can be detected by video endoscopy. The authors have already succeeded in synthesizing indocyanine green (ICG) derivatives for a fluorescent labeling substance which emits near-infrared rays. Before the antibodies labeled by these substances can be used, it is necessary to establish a method of vital immunohistochemical staining. So, we investigated the responses of antibodies exposed to non-fixed fresh tissue specimens as a basic study on vital immunohistochemical staining. The responses of fresh esophageal and gastric specimens (biopsied or surgically resected) to immunohistochemical staining with anti-epithelial membrane antigen (EMA) antibodies under various conditions using the ABC method were examined. These tissue specimens were stained immunohistochemically, and incubated with anti-EMA antibodies for 10 and 30 minutes (esophagus), and for 60 and 120 minutes (stomach) at 37 degrees C. These results suggest that vital immunohistochemical staining is possible under optimum conditions. If an infrared fluorescent endoscopy catching this excited fluorescence can be developed, it will be possible to establish a new endoscopic diagnostic method on the basis of vital immunohistochemical staining.

Labeled carcinoembryonic antigen antibodies excitable by infrared rays: a novel diagnostic method for micro cancers in the digestive tract

OBJECT

An indocyanine green derivative (ICG-sulfo-OSu) was used as the labeling substance for monoclonal antibody, and a fluorescence imaging system appropriate for ICG-sulfo-OSu excitable by infrared rays (IR) was developed. The goal of this study was to demonstrate antibody labeling at the tissue level using this new imaging system.

MATERIALS AND METHODS

ICG-sulfo-OSu labeled mouse anti-human carcinoembryonic antigen (CEA) monoclonal antibody, a newly developed imaging system, and an infrared ray microscope were employed in this experiment. Paraffin sections of human colon cancer previously proven to have cross-reactivity to anti-CEA antibody were examined.

RESULTS

Positive staining was seen as a brownish discoloration of oxidized 3,3'-diaminobenzidine tetrahydrochloride (DAB) in sections that reacted with ICG-sulfo-OSu-labeled anti-CEA antibody, and the fluorescence was well-matched with the oxidized DAB-positive sites.

CONCLUSION

Specific antibodies labeled with ICG-sulfo-OSu have significant affinity to cancer cells and seem to reflect sufficient amounts of fluorescence by IR to be useful in a system for the endoscopic detection of micro cancers using the immunohistochemical staining method.

Characterization of the heme and pyridoxal phosphate cofactors of human cystathionine beta-synthase reveals nonequivalent active sites

Cystathionine beta-synthase is an unusual enzyme that requires the cofactors heme and pyridoxal phosphate (PLP) to catalyze the condensation of homocysteine and serine to generate cystathionine. This transsulfuration reaction represents one of two major cellular routes for detoxification of homocysteine, which is a risk factor for atherosclerosis. While the beta-replacement reaction catalyzed by this enzyme suggests a role for the pyridoxal phosphate, the role of the heme is uncertain. In this study we have examined the effect of changing one of the ligands to the heme on the activity of the enzyme. Binding of carbon monooxide results in the displacement of a thiolate ligand to the ferrous heme, and is accompanied by complete loss of cystathionine beta-synthase activity. Furthermore, inhibition by CO is competitive with respect to homocysteine, providing the first indication that the homocysteine binding site is in the proximity of heme. Binding of both CO and cyanide to ferrous cystathionine beta-synthase occurs in two distinct isotherms and indicates that the hemes are nonequivalent. We have employed fluorescence spectroscopy to characterize the bound PLP and its interaction with serine. PLP bound to cystathionine beta-synthase is weakly fluorescent and exists as a mixture of the protonated and unprotonated tautomers. Reaction with hydroxylamine releases the oxime and greatly enhances the associated fluorescence. Binding of serine is accompanied by a shift to the unprotonated tautomer of the external aldimine as well as the appearance of a new fluorescent species at approximately 400 nm that could be due to the aminoacrylate or to a gemdiamine intermediate. These data provide the first characterization of the PLP bound to cystathionine beta-synthase. Treatment of cystathionine beta-synthase with hydroxylamine releases two PLPs after 1 day and results in complete loss of activity. Incubation for an additional 3-4 days results in the release of two more PLPs. These data lead us to revise the PLP stoichiometry to 4 per tetramer, and to the conclusion that the heme and PLP sites in cystathionine beta-synthase are nonequivalent.

Evidence for heme-mediated redox regulation of human cystathionine beta-synthase activity

Human cystathionine beta-synthase catalyzes the first step in the catabolic removal of the toxic metabolite, homocysteine. It is unique in being dependent on both pyridoxal phosphate (PLP) and heme for activity. The reaction involves condensation of serine and homocysteine to give cystathionine. Although the role of PLP can be rationalized in analogy with other PLP-dependent enzymes that catalyze beta-replacement reactions, the role of the heme is unknown. In this study, we have purified and characterized the recombinant human enzyme and have examined the effect of heme oxidation state on enzyme activity. We find that under reducing conditions, generated by addition of titanium citrate, the enzyme exhibits a 1.7-fold lower activity than under oxidizing conditions. Reoxidation of the ferrous enzyme with ferricyanide results in alleviation of inhibition. This redox-linked change in enzyme activity correlates with changes in heme oxidation state monitored by UV-visible spectroscopy. Dithiothreitol, which does not reduce the enzyme-bound heme, does not perturb enzyme activity. These studies provide the first evidence for redox-linked regulation of cystathionine beta-synthase which is heme-dependent.

Antibodies labeled with fluorescence-agent excitable by infrared rays

Endoscopy is not significantly better than fiberoscopy for the diagnosis of minute cancers of the digestive tract. However, labeling of these lesions with an agent that can be detected videoendoscopically, with subsequent computer processing of the electronic signals, should facilitate endoscopic diagnosis of microlesions. We developed an antibody labeled with an indocyanine green(ICG) derivative that has a specific fluorescence emission at 807 nm upon excitation at 768 nm. The physiochemical characteristics of this labeled antibody resemble those of ICG. The activity of the antibody is suitable for immunohistochemical staining, and the antibody fluoresces under infrared ray excitation. This antibody should prove useful for performing vital immunostaining for infrared endoscopy.

Development of agents for reinforcement of fluorescence on near-infrared ray excitation for immunohistological staining

Fluorescence intensity of indocyanine green (ICG) derivative (ICG-sulfo-OSu) was too low for its use to detect microlesions. Therefore, we examined the effects of reinforcement agents on ICG-sulfo-OSu labeled antibodies. Solutions of distearoylphosphatic acid sodium salt (DSPA) and octylglucoside (OG) in physiological phosphate buffered saline (PBS) were found to increase the intensity of fluorescence of ICG-sulfo-OSu labeled antibodies, with shift in the fluorescence peak wavelength from 804 to 821 nm.

Structural and electronic similarity but functional difference in methylmalonyl-CoA mutase between coenzyme B12 and the analog 2',5'-dideoxyadenosylcobalamin

The cofactor analog 2',5'-dideoxyadenosylcobalamin (ddAdoCbl) differs from the natural cofactor coenzyme B12 [5'-deoxyadenosylcobalamin (dAdoCbl)] by lacking only one oxygen atom. The 1H and 13C NMR spectra of ddAdoCbl have been assigned unambiguously by homonuclear and heteronuclear 2D NMR techniques. The 1H, 13C, and 31P chemical shift values for ddAdoCbl were compared with those of another organocobalamin, namely dAdoCbl. This assessment shows that the analog is very similar both electronically and structurally to the natural cofactor. The effectiveness of ddAdoCbl as a cofactor for both the human and Propionibacterium shermanii methylmalonyl-CoA mutases was compared with that of the natural cofactor. ddAdoCbl was found to be a competitive inhibitor with respect to dAdoCbl. Similar binding affinities to both enzymes were found for both the ddAdoCbl analog and the natural cofactor. However, in the presence of ddAdoCbl, the rate of conversion of methylmalonyl-CoA to succinyl-CoA was only 1-2% of that seen with the natural cofactor. There were no changes with time in the visible absorption spectrum of the bound cofactor analog in the presence of substrate, suggesting that the Co-C bond was not cleaved. The CD (circular dichroism) spectra of dAdoCbl and ddAdoCbl are very similar, consistent with the NMR results. The CD spectral changes upon binding to P. shermanii methylmalonyl-CoA mutase are large compared to those reported on the binding of dAdoCbl to ethanolamine ammonia lyase. Furthermore, the CD spectra of both enzyme-bound cobalamins are very similar, suggesting that similar changes in the conformation or structure in these cobalamins occur on binding to the enzyme.(ABSTRACT TRUNCATED AT 250 WORDS)

Inhibition of the human methylmalonyl-CoA mutase by various CoA-esters

Human methylmalonyl-CoA mutase is inhibited by ethylmalonyl-CoA, cyclopropylcarbonyl-CoA carboxylate, and methylenecyclopropylacetyl-CoA, which are substrate, intermediate, and product analogs, respectively. The mode of inhibition by each analog is reversible and mixed with respect to the substrate, methylmalonyl-CoA. This implies that the inhibitors are able to bind to both free enzyme and to the enzyme-substrate complex, although with affinities that are 4.5- to 10-fold different for the two species. The Ki1 for the cyclopropylcarbonyl-CoA carboxylate (0.26 +/- 0.07 mM), is 4-fold greater than the Km(app) measured for the substrate, methylmalonyl-CoA. Additionally, ethylmalonyl-CoA functions as an alternate substrate and is metabolized to methylsuccinyl-CoA. The human mutase is a homodimer that binds 1 mol of cobalamin per subunit. So, the observed mixed inhibition kinetics by substrate analogs is curious. Our finding that methylenecyclopropylacetyl-CoA, the causative agent of Jamaican "vomiting sickness," inhibits methylmalonyl-CoA mutase, while interesting, is probably not physiologically important because of the relatively high inhibition constants (Ki1 = 0.47 +/- 0.12 mM and Ki2 = 2 +/- 0.34 mM) observed with this compound.

Comparison of intra- and intermolecular proton transfer in human carbonic anhydrase II

The catalysis of the hydration of CO2 by human carbonic anhydrase II (HCA II) includes the transfer of a proton from zinc-bound water to histidine 64 utilizing a network of intervening hydrogen-bonded water molecules, then the proton is transferred to buffer in solution. We used stopped-flow spectrophotometry and 18O exchange between CO2 and water measured by mass spectrometry to compare catalytic constants dependent on proton transfer in HCA II and in the mutant H64A HCA II containing the replacement His64-->Ala. Maximal velocities and oxygen-18 exchange catalyzed by H64A HCA II showed that nearly all of the proton transfer with this mutant proceeded through the imidazole buffer. The following parameters were very similar or identical in catalysis by H64A HCA II compared with catalysis by wild-type HCA II both in the presence of large concentrations of imidazole (100 mM): the maximal rate of initial velocity and of exchange of 18O between CO2 and water, solvent hydrogen isotope effects on the maximal velocity, and the dependence of these isotope effects on the atom fraction of deuterium in solvent water. These results indicate that the proton transfer involving the zinc-bound water in catalysis is not significantly affected by the difference between the mobility of the free imidazole buffer and the side chain of His 64. Moreover, data for both the wild-type and mutant enzymes are consistent with proton transfer through intervening hydrogen-bonded water bridges in the active sites. These features of the proton transfer are discussed in terms of a model in which the first proton transfer from the zinc-bound water to an adjacent water is rate limiting.

Influence of amino acid replacement at position 198 on catalytic properties of zinc-bound water in human carbonic anhydrase III

Carbonic anhydrase III, found predominantly in skeletal muscle, is the least efficient of the mammalian carbonic anhydrases in catalyzing the hydration of CO2. Phenylalanine-198 is located on the hydrophobic side of the active-site cavity with its phenyl ring in the proximity of the catalytically active zinc-bound water. We replaced phenylalanine-198 in human carbonic anhydrase III with seven other amino acids (Ala, Asn, Asp, His, Leu, Tyr, Val) using site-directed mutagenesis. The catalytic properties of these enzymes were determined by stopped-flow spectrophotometry, and the exchange of 18O between CO2 and water was measured by mass spectrometry. All of the mutants had maximal values of kcat/Km for the hydration of CO2 enhanced, and five of the mutants had the pKa of the zinc-bound water increased compared with the wild-type enzyme. The largest effects were observed with the replacement Phe-198-->Asp which increased the maximal kcat/Km 140-fold and increased the pKa of the zinc-bound water from near 5 to 9.2. A Brønsted correlation was observed between log(kcat/Km) for hydration of CO2 and the pKa of the zinc-bound water (correlation coefficient r = 0.92); in addition, this pKa was inversely correlated with hydrophobicity of the residue at 198 (correlation coefficient r = -0.83). A direct correlation between the logarithm of the maximal kcat/Km for hydration and the logarithm of the pH-independent value of Ki for inhibition by cyanate (r = 0.95) indicated that the effect of the mutations at residue 198 occurred in large part by enhancement of the rate of dissociation of the enzyme-bicarbonate complex.

Slow oxygen release on the first two flashes in chemically stressed Photosystem II membrane fragments results from hydrogen peroxide oxidation

Flash-induced amperometric signals were measured with a Joliot-type O2 rate electrode in spinach Photosystem II (PS II) membrane fragments exposed to very low concentrations of added hydroxylamine or hydrogen peroxide. In both cases 'anomalous O2 signals' were observed on the first two flashes, and oscillating four-flash patterns were observed on subsequent flashes. The anomalous signals were eliminated in the presence of catalase but not EDTA. The rise times of the O2-release kinetics associated with the anomalous signals were slow (ca. 20 ms with NH2OH and ca. 120 ms with H2O2) compared to the kinetics of O2 release on subsequent flashes and in control membranes (3-6 ms). It is proposed that when the intact PS II O2-evolving complex is perturbed with small concentrations of added reductant, H2O2 can gain access and bind to the complex. Bound H2O2 can then reduce lower S states in some centers leading to anomalous O2 signals on the first two flashes. A model is presented to explain both types of anomalous O2 production. Oxygen observed on the third and subsequent flashes is due to the normal photosynthetic O2-evolution process arising from the S3-state. Anomalous O2 production could be a protective mechanism in PS II centers subjected to stress conditions.

Catalysis by mutants of human carbonic anhydrase II: effects of replacing hydrophobic residues 198 and 204

Previous studies shows that the replacement of Phe-198 in carbonic anhydrase III to the corresponding Leu residue found in carbonic anhydrase II caused the appearance of isozyme II-like activity (LoGrasso et al. (1991) Biochemistry 30, 8463-8470). Carbonic anhydrase II is more efficient in the catalysis of CO2 hydration by 500-fold and has an apparent pKa for this catalysis about two pKa units above that of carbonic anhydrase III. Moreover, isozyme II catalyzes the hydrolysis of 4-nitrophenyl acetate, whereas isozyme III shows no appreciable catalysis. The purpose of this work was to test the hypothesis that making the converse replacement Leu-198-->Phe as well as Leu-204-->Glu and the double replacement in carbonic anhydrase II would give the resulting mutants of isozyme II properties of isozyme III. The catalytic activities of these mutants in CO2 hydration and 4-nitrophenyl acetate hydrolysis were smaller by at most 5-fold and the pKa values for these catalyses were identical compared with wild-type isozyme II. The different effects of converse mutants of HCA II and III indicate complexity in structure not evident from their similar backbone conformations.

Adenosine deaminase and purine nucleoside phosphorylase activities in lymphocytes from patients with lung cancer

The activity of adenosine deaminase (EC 3.5.4.4.) was significantly lower in lymphocytes from patients with lung cancer than in those from healthy subjects, whereas the activity of purine nucleoside phosphorylase (EC 2.4.2.1.) was significantly higher in lymphocytes from the patients than in those from normal controls. When the one activity was plotted against the other, the plots for patients with lung cancer were all outside the frame formed by the lower and higher limits of the standard deviation of the mean of normal activities of the two enzymes. The ratio of adenosine deaminase activity to purine nucleoside phosphorylase activity was lower in patients with lung cancer than in controls. The possible effect of this ratio on the function of lymphocytes was briefly discussed. These enzyme activities were suggested to be useful measures of the immune responsiveness of patients with lung cancer.

Phytohemagglutinin skin test: diagnostic value for showing immunodeficiency in patients with cancer

The skin response to 5 mug of purified phytohemagglutinin (PHA) was studied in 299 subjects, including 58 normal controls, 92 patients without malignancies, and 149 patients with nonlymphomatous cancer. Other immunological responses, such as in vitro lymphocyte stimulation (62 subjects) and skin response to purified protein derivatives (PPD) (95 subjects), were tested simultaneously to examine their correlation with the PHA skin test. A positive reaction was observed 24 hr after intradermal injection of 5 mug of purified PHA in 56 (96.6%) of 58 normal controls, 40 (49.4%) of 81 untreated patients with cancer, and 24 (35.3%) of 68 cancer patients receiving anticancer therapy. Among 32 patients with gastric cancer tested, impaired skin reactivity to purified PHA was noted in patients in stage III or IV. A correlation was found between in vivo and in vitro responses to PHA in 46 (74.2%) of 62 individuals (P less than 0.001). The PHA skin test was repeated 4 times over a period of three months in patients without malignancies, and no significant change in tehir skin reactivity was detected. In repeated tests, the skin reactivity to purified PHA of patients with lung cancer varied depending on the clinical status, and the extent and type of anticancer therapy the patients were receiving. It is concluded that the PHA skin test is a simple diagnostic method for screening for immunodeficiency in cancer patients before and during the course of anticancer therapy. Other advantages of this test are that no presensitization is required and that it can be used repeatedly.