Chediak-Higashi Syndrome Neutrophilsare Characterized

Home , Azurophilic granule, Phagolysosome

Chediak-Higashi Syndrome Neutrophils Are Characterized by the A bsence of Both Normal Azurophilic Granules

BURION C. WEST, MD From the Section of Infectious Diseases, Department of Medicine, Louisiana State University School of Medicine, Shreveport, Louisiana; and the Laboratory of Clinical Investigation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland

Neutrophils from two Chediak-Higashi syndrome identified was eosinophil granule peroxidase at Q = 1.24 brothers were isolated, suspended in heparinized sucrose, g/ml (band E). Alkaline phosphatase, not a granule lysed, and filtered. The granule-rich filtrate was cen- marker, was twice normal at the normal density, Q = trifuged on a sucrose gradient (Q = 1.287-1.10 g/ml) at 1.14-1.15 g/ml, consistent with an increase in unidentified a mean force of95,000g for 4 hours. The gradients con- membranes. A lysate gradient suggested that the giant tained one band at Q = 1.18 g/ml (band C) which was azurophilic granules were Q = 1.25-1.27 g/ml. These neu- broader than normal and lacked normal bands A, Q = trophils contain blue-grey or slate-grey giant granules, 1.22 g/ml, and B, Q = 1.20 g/ml. Gradient fractions were which are not truly azurophilic or basophilic, but should assayed for enzyme activities and protein. No marker en- continue to be identified as azurophilic to conform to the zymes identified densities normally occupied by bands convention making "azurophilic" and "peroxidase- A and B, and of the enzymes measured, only lysozyme positive" synonymous. The eosinophils contain normal showed peak activity with band C. Thus, only normal eosinophil granules as well as giant inclusion granules. specific granules were present. Two azurophil gran- In contrast, neutrophils are deficient in both normal ules, normally present and separable, were absent. Also azurophilic granules. (Am J Pathol 1986, 122:177-189)

THE CHEDIAK-HIGASHI SYNDROME, described ules, which appear to form from fusion of azurophil in 1943,1 is a rare autosomal recessive disorder, fre- and specific granules.2 8 12 Moreover, no studies address quently complicated by infections, occurring in man the number and character of granules present in the pe- and animals; it is regarded as a lysosomal disease be- ripheral blood neutrophil of human beings with the cause of the presence of giant granules with lysosomal Chediak-Higashi syndrome. The studies presented here properties in leukocytes and other cells.2-4 Accessible address this question. They provide no evidence for the neutrophils from human beings with the Chediak- presence of normal azurophil granules of any kind in Higashi syndrome have been extensively studied,5-18 al- the Chediak-Higashi syndrome neutrophil from periph- though few studies have examined isolated neutrophil eral human blood. Furthermore, they support the con- granules.7'13-18 Interest in the formation of the giant cept that all azurophil granules in this cell are in fact granules in neutrophils has been histologic and cyto- abnormal. They show that specific granules are pres- chemical.2' 1015 Even the deficiency of natural killer cells ent at the density in sucrose gradients at which normal in the Chediak-Higashi syndrome appears to be related specific granules are known to be at equilibrium to reduced numbers and defective functioning of large density25; this previous report of granule separation granular lymphocytes19 where poorly characterized large using normal human neutrophils constitutes concur- abnormal granules might constitute the major defect. rently performed controls for this report.25 The method Concurrent interest has focused upon neutrophil mem- of granule separation has been confirmed.26-31 branes, particularly the membrane lipids and micro- tubular interconnections of membranes.20-24 Membrane Accepted for publication August 16, 1985. and microtubule defects are present, and although not Address reprint requests and correspondence to Burton C. completely characterized, have been considered impor- West, MD, LSU Medical Center, P.O. Box 33932, Shreveport, tant to understanding the formation of the giant gran- LA 71130-3932.

177 178 WEST AJP * January 1986 Materials and Methods ameter, often exceeding 2 .. 4,12,15,34,35 In two experi- Patients ments 0.1-ml aliquots of lysates and 5-,u filtrates, and in one of these experiments, 0.1 ml of the 5-,A and 2-,p Two brothers with the Chediak-Higashi syndrome2.32 filtrates, were diluted to 2.0 ml with 0.34 M sucrose con- were the source of blood neutrophils. At the time of taining 0.25 mM Na2H2EDTA and assayed for enzyme these experiments they were 22 and 21 years old and activities and protein (Table 1). In another, a lysate and were free of infection. Concurrent parallel control ex- a 5-si filtrate were separately subjected to density gra- periments using venous blood from 14 normal adult dient centrifugation. donors as the source of neutrophils previously reported provided the description of normal neutrophil granules.25'26 These data are not repeated here. Permission Sucrose Density Gradients after informed consent for venipuncture was obtained sucrose gradients were made as previ- with guidelines. Continuous in keeping institutional ously described with the use of a 3-ml cushion of sucrose, Q = 1.287 g/ml, 15 ml of the same dense sucrose, Materials and 15 ml of light sucrose, Q = 1.10 g/ml. Volumes were to allow room for a 2.5-5.0-ml sample at the chemicals were used throughout. Wa- adjusted Reagent grade top.25 ter was triply distilled and autoclaved. For centrifugation we used a Model L2-65B and an SW 27 rotor with cups which held 38.5 ml cellulose ni- Blood Neutrophils trate tubes (Beckman Instruments Inc., Palo Alto, Calif). Centrifugations were conducted at 4 C for 4 Peripheral venous blood was drawn in 5 mM hours at 27,000 rpm (mean force, 95,000g). Gradients Na2H2EDTA, pH 7.3, in a volume of 100 ml twice from were photographed and collected from below into 41-46 La. R. and 200 ml once from Le. R. Neutrophils were fractions and the pellet fraction, which was scraped promptly isolated by Hypaque-Ficoll gradients and dex- from the bottom and resuspended in one ml 0.34 M tran sedimentation and clarified of erythrocytes.25 33 The sucrose.25 Fractions were frozen at -20 C and assayed leukocytes, greatly enriched in neutrophils, were cen- promptly, all within 10 weeks. trifuged at 150g and washed in 30 ml of 0.34 M sucrose with 0.25 mM Na2H2EDTA; then the cell suspension was centrifuged at 150g and adjusted to 50 x 106 leu- Biochemical Determinations kocytes/4 ml of 0.34 M sucrose with 0.25 mM Na2H2EDTA. A Gilford Model 240 spectrophotometer (Gilford In- strument Laboratories, Inc., Oberlin, Ohio) was used throughout. Because of fraction volume, all assays could Preparation of Neutrophil Granules not be performed on every gradient. The number of To each 4-ml cell suspension was added sodium hepa- gradients on which an assay was performed is indicated. rin, 0.5 ml, containing 5000 units (The Upjohn Com- Assay descriptions and conditions are identical to those pany, Kalamazoo, Mich). Cell disruption, aided by previously reported.25 Protein, utilizing egg white lyso- heparin, was accomplished by 11-15 cycles of suction- zyme as standard, is expressed as Mg/ml.36 Because of ing the suspension through a 9-cm 18-gauge needle at- interference of hydrolyzed sucrose in the accuracy of tached to a 6-ml polypropylene syringe.2526 Because of the Lowry method, the values reported are considered concern about the destruction of the giant azurophilic estimates.37 Myeloperoxidase (EC 1.11.1.7) is expressed granules by the lysis procedure, an attempt was made as microgram equivalents of horseradish peroxidase per to minimize resuctioning. Lysis was monitored by phase milliliter.25 38 Lysozyme (EC 3.2.1.17) is expressed as microscopy. micrograms of egg white lysozyme equivalent per mil- Lysates were filtered through a 5-,. (pore size) poly- liliter.39 (-Glucuronidase (EC 3.2.1.31) is expressed as carbonate filter, 25 mm in diameter (Nuclepore, General micrograms phenolphthalein released per hour per mil- Electric Corp., Pleasanton, Calif), followed by filtra- liliter.40'41 Acid phosphatase (EC 3.1.3.2) was determined tion of the 5-,A filtrate through a 2-,u pore size filter.25 by two methods and is expressed either as micromoles Each 5-. and 2-, filtrate was applied to a sucrose gra- of p-nitrophenylphosphate released per hour per mil- dient. liliter or as nanograms inorganic phosphorous liber- A limited effort was made to assess the presumably ated (from sodium P-glycerophosphate) per minute per deleterious effects of filtration of Chediak-Higashi syn- milliliter.43'44 Alkaline phosphatase (EC 3.1.3.1) is ex- drome giant granules, which range from 1 to 4 IA in di- pressed as micromoles of p-nitrophenylphosphate Vol. 122 * No. 1 CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 179

Table 1-The Distribution of Protein Concentration and Enzyme Activities in Lysates, 5-p (Pore Size) Filtrates, and 5-J and 2-M (Pore Size) Filtrates of Normal Human Neutrophils and Ch6diak-Higashi Syndrome Neutrophils, the Latter as Estimated From Two Experiments Normals* P Ch6diak-Higashi syndrome Protein (mg/ml) Lysate 0.85 ± 0.03 (22)t <0.01 0.45, 0.52 5-f filtrate 0.62 ± 0.04 (21) NS 0.45, 0.59 5-2-1j filtrate 0.62 ± 0.02 (22) NS 0.63, 0.48 Enzyme activityt Myeloperoxidase Lysate 0.94 + 0.05 (23) NS 0.85 ± 0.16 (4) 5-r filtrate 0.77 ± 0.04 (23) <0.001 0.33 ± 0.04 (4) 5-2-MA filtrate 0.76 ± 0.04 (22) <0.001 0.23, 0.24 P-Glucuronidase Lysate 74.6 ± 5.6 (11) NS 55 5-p filtrate 55.6 ± 6.5 (10) <0.05 15 5-2-IA filtrate 63.2 ± 5.4 (10) - - Lysozyme Lysate 72.1 + 4.5 (18) <0.001 137 ± 40 (3) 5-,. filtrate 54.9 ± 4.2 (18) NS 53, 47 5-2-M filtrate 55.8 ± 3.6 (17) NS 62, 73 Alkaline p-nitrophenylphosphatase* Lysate 0.88 ± 0.05 (22) <0.001 2.70 ± 0.35 (4) 5-;* filtrate 0.63 ± 0.05 (21) <0.001 1.23 ± 0.05 (4) 5-2-;A filtrate 0.66 ± 0.05 (21) <0.001 1.67, 1.67 Acid p-nitrophenylphosphatase Lysate 4.49 ± 0.38 (23) <0.001 10.7, 11.0 5-IA filtrate 3.13 ± 0.28 (23) NS 1.6, 1.5 5-2-IA filtrate 2.81 + 0.24 (21) Acid p-glycerophosphataset Lysate 0.54 ± 0.08 (3) NS 0.74 ± 0.12 (3) 5-JA filtrate 0.48 + 0.03 (3) <0.01 0.28 ± 0.03 (4) 5-2-uA filtrate 0.46 + 0.09 (3) NS 0.20, 0.20 Normal data (column A) are directly from West et al,26 except alkaline phosphatase, which was corrected by multiplying the previously reported data by 0.5. t More than two data are expressed as the mean ± SEM; n is in parentheses. t Enzyme activity is expressed as enzyme units (see Materials and Methods) per milliliter of lysate or filtrate, except that acid P-glycerophosphatase is expressed as micrograms of phosphorus released per minute per milliliter. released per hour per milliliter.45 Enzyme activities were Means are compared with the t test for paired data linear under the experimental conditions. or for unpaired data. The null hypothesis was rejected Heparin inhibition of P-glucuronidase and acid p- for P < 0.05. Programs for these calculations and de- nitrophenylphosphatase was reversed by incorporating termination of linear regression were used in a program- 0.1 ml of protamine sulfate (1 mg/ml, Upjohn) in these mable calculator (HP97, Hewlett-Packard Co., Cor- reaction mixtures. 15.16.25 Acid P-glycerophosphatase is vallis, Ore). mildly inhibited by heparin.46 The effect is not easily overcome with protamine, and therefore the values are considered estimates. 16.25.26 Results Neutrophil Separation and Lysis Photomicrographs The patients were neutropenic, the mean white blood Wright's-stained blood smears were photographed cell count being 4775 ± 550414 (± SD), with 31.107o with a Nikon microscope camera using color print film ± 5.8% neutrophils (absolute neutrophil count, 1463 (Vericolor III, Eastman Kodak Co., Rochester, NY). ± 15241ll), 63.5% ± 4.3% lymphocytes, 3.17o ± 2.1% monocytes, and 2.77o ± 1.2% eosinophils. The per- to Statistical centage of neutrophils after preparation rose 94.1% Analysis ± 1.9%. In addition, the leukocytes that were lysed con- Data are expressed as the mean ± standard devia- tained 2.1% ± 1.2% lymphocytes and 3.8% ± 0.7% tion (SD) (SEM, Table 1) for an estimate of variation. eosinophils. The recovery rate of neutrophils was 63 7o, 180 WEST AJP * January 1986

the 0.34 M sucrose used to suspend and wash neutrophils. After 8-10 cycles of resuctioning, some unlysed neutrophils were evident by phase-contrast microscopy. Granule suspensions were handled gently for minimiz- ing damage to the giant granules. After additional lysis resuctioning totaling 11-15 cycles, the lysate displayed an impressive number of generally round, giant granules, approximately 1-4 , in diameter, in monodispersed suspension, and many smaller, normal-sized granules, not distinguished one from another. A few unlysed neutrophils, rare erythrocytes, and many neutrophil nuclei were also present.

Comparison of the Neutrophil Lysate With Filtrates Passage of the lysate through the 5-M (pore size) filter was as difficult as in normals, with some resistance, but even Figure 1-A representative gradient of monodispersed neutrophil gran- flow. Passage through the 2-I (pore size) filter was ules from a patient with the Ch6diak-Higashi syndrome is located on the more difficult, with frank resistance. By phase micros- right, demonstrating one dense visible band, band C, at Q = 1.18 g/ml, copy the 5-,u (pore size) filtrate showed an occasional and an extremely faint band E at Q = 1.24 g/ml. For comparison, the left panel is a photograph of a gradient of monodispersed neutrophil granules Chediak-Higashi syndrome giant granule. Very rare from a normal adult, which demonstrates the existence of three dense appearance ap- bands, A, B, and C. Bands A and B are located at densities 1.22 g/ml and granules slightly larger than normal in 1.20 g/ml and contain normal azurophilic (peroxidase-positive) granules; peared in the 5-* and 2-si filtrate, consistent with the such bands and the granules they represent are absent from Ch6diak- Higashi syndrome neutrophils. absence of giant granules from this filtrate and from subsequent isopyknic sucrose density gradients. A comparison of the enzyme activities of lysate with which was the mean of yields 59% and 57% on cell each of the two filtrates in one experiment and a com- separations for La. R. and 72% on the one for Le. R. parison between the lysate and the 5-. (pore size) filtrate in a second constitute a limited assessment of the effects Neutrophil Granules of filtration. In the latter, no 5-M filtrate could be forced through the 2-, filter because it was so thick and vis- Following addition of heparin to neutrophils in 0.34 cous; hence, no 5-1. and 2-. filtrate was available. No M sucrose, cells were observed by phase microscopy to change in protein concentration was caused by filtra- undergo slow spontaneous lysis with a gradual increase tion (Table 1). The concentration of enzyme activity for in free granules in suspension. Previously, addition of peroxidase and the acid hydrolases, acid P-glycero- 1 mM CaCl2 to such suspensions prevented neutrophil phosphatase and P-glucuronidase (limited to the 5-,u lysis and the liberation of granules, and the neutrophils filtrate), declined precipitously from approximating nor- were shown to have many projections from the cell sur- mal levels in lysates, to levels much lower than normals face.25 Because EDTA does not interfere in neutrophil after filtration.26 Lysate lysozyme was nearly twice nor- lysis, and millimolar amounts of CaCl2 have adverse mal (137 versus 72 U/ml) and decreased to levels com- effects upon the neutrophils, their lysis, and liberation parable to normal after filtration through the 5-,I filter, of free granules, 0.25 mM Na2H2EDTA was added to compatible with destruction of the giant granules, leav-

Figure 2-Protein and enzyme activity of normal (left, IA, IIA, and IIIA reproduced from West et al,25 Text-Figure 2) and Ch6diak-Higashi syndrome neutrophil granules (right, IB, IIB, and IIIB) in sucrose gradients from density 1.287 g/ml to 1.10 g/ml. Gradient bands are represented by the arrows below the letters designating the bands. Band A and band B are present only in normals (left), while band C (see Figure 1) is present in gradients from both normal and Ch6diak-Higashi syndrome neutrophil granules. I-Myeloperoxidase activity and protein. IA-Myeloperoxidase (mean of 14 gradients), line a, solid; protein (mean of 11 gradients, milligrams per milliliter), line b, dashed. IB-Myeloperoxidase (mean of 3 gradients), line a, solid; protein (mean of 3 gradients), line b, dashed. Levels are shown. II-p-Glucuronidase and lysozyme activities. IIA-p-Glucuronidase (mean of 6 gradients), line c, solid; lysozyme (muramidase) (mean of 8 gradients), line d, dashed. IIB-p-glucuronidase (one gradient), line c, solid; lysozyme (mean of 3 gradients), line d, dashed. Levels are shown. III-Phosphatase activities. IIIA-Acid P-glycerophosphatase (mean of 4 gradients), line e, dashed (ordinate on right); alkaline p-nitrophenylphosphatase (mean of 14 gradients), line f, solid; acid p-nitrophenylphosphatase (mean of 13 gradients), line g, short dashes. IIIB-Acid P-glycerophosphatase (mean of 2 gradients), line e, dashed (ordinate on right); alkaline p- nitrophenylphosphatase (mean of 2 gradients), line f, solid; and acid p-nitrophenylphosphatase (one gradient), line g, dotted. Levels are shown. Alkaline phosphatase activity (IIIA and IIIB) should be multiplied by 0.025 to yield activity in micromoles per hour per milliliter. See text (Materials and Methods) for enzyme units. Vol. 122 * No. I CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 181

024 01 a' 0.20

0.16 B 0.2 1 T - 0.16rAoMPO 110.12 , jb Protein 0160 1\ Ib~~~~~~~~~~~lbProtein c i 0.08~~~~~~~~~~~~~~~00

0.04 - - 0.02 - MP 0 5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45 e Q1.287 FRACTION 1.10 FRACTION A B C 10.0 c 3-Glucuronidose 6 4 8.0 -Td Lysoyree ~6.0 \- ~~~~\4I~~~~~~~~5T ~dLyswZfrri V1~~~~~~~~~~~~I 4.0 <~~~~~~~~~~~~~~~-'3' 1.0- c/ i6r" 1.0- A -*,.-~~~~~~~~~~~~VAnidase' (0 5 10 1520 25 3035 4045 50 5 20 25 30 3'54045 Q1.287 FRAMON ~~~~Ql.10OU PrTM II

FRACTION 182 WEST AJP * January 1986 ing normal specific granules in the filtrate. Alkaline others in this peak. When each pair of these means was phosphatase was decreased by filtration, but less so than compared with the t statistic for two means, there was the acid hydrolases. These results are approximations no significant difference. Nevertheless, these values were because of limited data and of 20-fold dilution of ly- similar to the peak lysozyme activity for band C in an sates and filtrates. earlier report of 5.6 ± 1.5 units,31 the activity in Fraction 23 of the original report of comparable gradients of Sucrose Gradients 8.9 ± 2.1 units, and the activity in Fraction 22 of the broader band C lysozyme peak of "expanded" gra- Figure 1 presents a representative gradient of gran- dients of 4.3 ± 1.9 units.25 No current mean peak ules from neutrophils of a patient with the Ch- lysozyme value is different from the previously pub- diak-Higashi syndrome, compared with normal. In lished ones cited. However, the width and intensity of both, 50 x 106 neutrophils were the source of the gran- the Chediak-Higashi syndrome neutrophil granule ules. The comparison highlights the presence of band band C are greater than the width and intensity of band C in the Chediak-Higashi syndrome neutrophil gran- C in equivalent normal granule gradients (Figure 1).25 ule gradient, located in exactly the same density as band Thus, there was a contrast between normal lysozyme C in normal neutrophil granule gradients (Q = 1.18 peak levels in band C, which suggests a normal com- g/ml). It is more intense in light scattering and is wider plement of "specific" granules in Chediak-Higashi neu- than is a normal band C. No band A or band B is pres- trophils, and the greater physical width and intensity ent in any gradient of granules from the Chediak- of band C from these patients, which might be com- Higashi syndrome neutrophils. At Q = 1.24 g/ml, a very prised of other granules of Q = 1.18 g/ml, not contain- faint band E was identified. Figure 2 shows the gradient ing lysozyme. Because no other measured enzyme was distributon of protein and six enzymes in granules sub- present, giant granules are excluded at Q = 1.18 g/ml. jected to isopyknic density centrifugation. Band C is Alkaline phosphatase peak activity was increased, indicated for Fraction 23, which was the midpoint of compared with normals, and was situated in an other- the only major band seen in these gradients (Figure 1) wise enzyme-"deficient" region, not associated with (e = 1.18 g/ml). The single peak of lysozyme activity granules (Figure 2). Its location in a broad peak (Q = (Figure 2) corresponded to band C. Note the very small 1.14-1.15 g/ml) was identical to gradients of normal peak of isolated peroxidase activity at Fraction 13; it granules.25 correlates to the few eosinophil granules known to sedi- Another means of examining the distribution of the ment to Q = 1.24 g/ml and is identified as band E.2 526 enzyme activity is to tally the total activity for a gra- In the present study, band E was barely visible, reflect- dient by adding the units of enzyme activity (or pro- ing the small percentage of eosinophils of the leuko- tein concentration) for all fractions. Total activity is used cytes lysed (2.7% ± 1.2%). In contrast to normal neu- as a denominator for the sum of selected fractions where trophil granules, there was not even the suggestion of activity is concentrated. Such data are compiled and activity other than background at Q = 1.22 g/ml (Frac- grouped in Table 2. The average activity per tube within tion 16) and Q = 1.20 g/ml (Fraction 20), sites of bands the data group which represents a peak is also expressed. A and B in equivalent gradients of normal neutrophil These data confirm peaks near the origin, which reflect granules.25 solubilization of lysosomal enzymes by the lysis or filtration procedure and estimations created because of the use of heparin and protamine. Acidp-nitrophenylphos- and Total Activities Peak Activities Gradient phatase is generally associated with protein, a non- The midpoint of the peak of lysozyme activity was lysosomal location. The only true peak within the Fraction 23, marked "C" on Figure 2, which correlated gradient which was granule-associated was that of lyso- to band C at Q = 1.18 g/ml (Figure 1); equivalent gra- zyme. Alkaline phosphatase was localized to Fractions dients of normal granules showed the same band.25"3 29-42 (Table 2) and was not associated with other en- Mean activity in Fraction 23 was less than in adjacent zymes or with identified granules. fractions. The absence of a true peak at Fraction 23 with appeared to be due to an unexplained problem the Sucrose Gradient of Unfiltered Lysate lysozyme determination in one gradient, which yielded much less activity than Fraction 22 or 24. The lysozyme The gradient of the lysate of 50 x 106 Chediak- (mean ± SD) for fractions in this peak were as follow: Higashi syndrome leukocytes, containing 960/o neu- Fraction 22, 6.3 ± 0.9 units; Fraction 23, 5.30 ± 2.94 trophils, 10o lymphocytes, and 3% eosinophils, was units; and Fraction 24, 5.9 ± 0.5 units. The standard collected into 39 fractions and a pellet fraction and deviation for Fraction 23 was large, compared with contained band C and faintly visible band E (Figure Vol. 122 * No. 1 CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 183

Table 2-Percentage of Protein and Enzyme Activity Distributed in Designated Portions of the Gradients, Identified by Inclusive Gradient Fraction Numbers, and the Average Protein Content or Enzyme Activity per Fraction in That Designated Group of Fractions, in Parentheses, Which Provides a Comparison of One Peak or Possible Peak With Another Pellet Test Pellet and 1-17 19-27 29-42 38-46 Protein (3)* 3.0 (77)t 26 (35) 12 (35) 32 (59) 51 (146)t Myeloperoxidase (3) 2.5 (9) 18 (4) 11 (4) 45 (12) 73 (30) P-Glucuronidase (1) 2.2 (0.5) 22 (0.3) 18 (0.5) 43 (0.8) 45 (1.8) Lysozyme (3) 1.6 (1.4) 8.7 (0.4) 39 (4.3) 27 (1.8) 32 (3.3) Alkaline p-nitrophenylphosphatase§ (2) 0.2 (0.3) 0.9 (0.1) 4.7 (0.9) 84 (10.5) 37 (7.2) Acid p-nitrophenylphosphatase (1) 0.8 (0.4) 2.7 (0.1) 17 (1.0) 58 (2.4) 41 (3.1) Acid p-glycerophosphatase (2) 0.7 (3.7) 4.2 (1.3) 18 (10.3) 45 (17) 59 (35) * Numbers in parentheses after the test name indicate gradients assayed. t Numbers in parentheses in each row are the average protein content (micrograms per milliliter) or enzyme activity (see text for units) per fraction for the designated group or possible peak. t Underlined values represent peaks. § Activity should be multiplied by 0.025 to yield micromoles per hour per milliliter.

3, left). Neither band A nor band B was present. This lysozyme with peak activity of 11.1 units in Fraction gradient at and above band E appeared similar to a 7. At this peak, the specific activity is 72 U/mg protein gradient of the 5-s and 2-j filtrate run simultaneously and is comparable to that observed in normal neutrophil (Figure 3, right). Below band E, however, was a dense lysate (83 U/mg protein).26 band extending into the cushion to within 8 mm of the These results indicate that gradient centrifugation bottom. The upper one-third of this 12-mm band con- caused a significant specific activity of both myeloperox- tained numerous particles or clumps up to 3 mm in di- idase and lysozyme to be at Q = 1.25-1.27 g/ml in as- ameter. Some appeared connected to strands which ex- sociation with a broad visible band. The experiments tended down; however, the lower two-thirds was free do not identify giant azurophilic granules at Q = of the larger, particlelike structures. Because of these 1.25-1.27 g/ml but are highly suggestive of their pres- proteinaceous strands and particles, the gradient was ence. They appear to share this density with other pro- not smoothly collected; the gradient contents were dis- rupted vertically by strands pulled into the collecting needle through several levels. Caution is required in interpreting results of gradients containing these proteinaceous strands. Protein levels defined this broad band from Fraction 2 to Fraction 11, each fraction containing more than 55 *g/ml. The peak in Fraction 7 contained 153 ,g/ml; at higher density below this peak was a prominent shoulder. A nearly identical peak of myeloperoxidase activity extended from Fraction 2 (0.035 units) through Fraction 11 (0.032 units). The peak at Fraction 7 contained 0.209 units. Therefore, the specific activity of myeloperoxidase at Fraction 7 (1.37 U/mg protein) is higher than that of normal neutrophil lysates (1.09 U/mg protein) and yet does not approach the specific activity of 7.07 U/mg protein, which is the peak occurring in the less dense normal azurophil granules in band B at Q = 1.20 g/ml.25 Lysozyme and alkaline phosphatase were determined on this gradient, but acid hydrolases were not. Lyso- zyme has its peak at Fraction 17, Figure 3-Photographs of three gradients representing the Chediak- which is band C in Higashi syndrome neutrophil granules. Left-Lysate gradient. this gradient; its activity is 6.8 units and specific activ- Center-Five-micron filtrate gradient. Right-Five- and 2-p filtrate gra- ity 179 U/mg protein. This value compares to dient. All three wero performed on one blood sample and simultaneously normal centrifuged. Band C and band E are demonstrated (see Figure 1 and text). lysate specific activity of 83 U/mg protein26 and a band The broad band at density range Q = 1.25-1.27 g/ml lying under band C peak in normal gradients of 342 U/mg protein.25 The E in the lysate gradient (left) and the faint broad band in the same location in the 5-;A filtrate gradient (center) probably represent the giant 12-mm band below band E also had a broad peak of azurophilic granules of the Ch6diak-Higashi syndrome neutrophil. 184 WEST AJP * January 1986

A B

Figure 4-Photomicrograph of a Ch6diak-Higashi syndrome neutrophil (A) and an eosinophil (B) from peripheral blood stained by the Romanowsky method with Wright's stain, demonstrating giant inclusion granules in both, but cytoplasmic granules typical of the cell only in the eosinophil. (Original magnification, x 1000)

teins. Although these enzyme activities are not more A broad peak of alkaline phosphatase was observed concentrated than lysate, considering their location in in Fractions 25-34, with a peak in Fraction 27 of 9.8 the gradient with a broad visible band and the absence units and specific activity of 316 U/mg protein. This of a similar band and enzymes from comparable gra- compares favorably with the lysate specific activity in dients of neutrophil lysates from normal human beings, normals of 2.1 U/mg protein.26 This concentration of it is reasonable to think they represent giant azurophilic activity is also comparable to the peak in gradient Frac- granules characteristic of Chediak-Higashi syndrome tion 35 of Chediak-Higashi syndrome 5-,u and 2-,. neutrophils. filtrates of 12.8 units and specific activity of 582 U/mg Vol. 122 * No. 1 CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 185 protein. In contrast to gradients of the Chediak-Higashi of giant azurophilic inclusions of the eosinophil and syndrome 5-,A and 2-. filtrates and to gradients of nor- of true eosinophil granules is more distinct than the sub- mal neutrophil filtrates, the unique 12-mm band below tle but real difference in color between the giant band E contained alkaline phosphatase, with peak ac- azurophilic granule in Chediak-Higashi syndrome neu- tivity (Fraction 7) of 24.9 units and specific activity of trophils and the color of azurophilic granules in nor- 163 U/mg protein. This is considerably less than peak mal neutrophils. Normal azurophilic granules are, in specific activity at fraction 27 in the same gradient, the fact, lighter and paler than giant azurophilic granules Chediak-Higashi syndrome 5-,u and 2-, filtrate gradient but cannot be shown on the same photograph because peak at Q = 1.14-1.16 g/ml, and the normal neutrophil they do not appear to exist. alkaline phosphatase peak seen at Q = 1.145 g/ml of 433 U/mg protein.25 Discussion These experiments show that the Chediak-Higashi Sucrose Gradient of 5-, Filtrate syndrome is characterized by the absence of both nor- With additional lysate prepared in the preceding ex- mal azurophil granules. This deficiency is central to periment, the 5-. (pore size) filtrate was prepared and what is wrong with these persons, probably being equal layered on an identical sucrose density gradient and cen- to if not more important than the existence of the dra- tifuged with the gradients described above (Figure 3). matic giant azurophilic granules which have captured The resulting isopyknic gradient was identical to the so much attention. The other surprising result of these lysate gradient, except that the band containing globu- studies is that the giant granules appear to have a den- lar particles below band E was less intense and the sity of 1.25-1.27 g/ml in preliminary experiments, which strands of proteinaceous material running vertically be- constitutes a density greater than that of all known nor- low that density were considerably less evident (Figure mal neutrophil or eosinophil granules. 3, center). The gradient was collected into 43 fractions The question of whether or not Chediak-Higashi and a pellet fraction. The peak of myeloperoxidase and syndrome neutrophils in human beings contain normal protein (Q = 1.25-1.27 g/ml) was less than the lysate azurophilic granules has been addressed by histochem- gradient and more than in the 5-. and 2-p. filtrate gra- ical and cytochemical techniques. The presence of dients described in detail, where this broad band with normal azurophilic granules coexisting with giant associated myeloperoxidase activity and protein does azurophilic granules has been asserted, although a very not exist. The broad peak extended from Fraction 4 to small number of profiles possibly consistent with nor- Fraction 20 with a maximum myeloperoxidase activity mal azurophilic granules have been demonstrated only in Fraction 11 of 0.064 units (protein, 35 ,g/ml). Other in certain patients.4',2 A different interpretation could enzymes were comparably intermediate between the be attached to the published transmission electron other two types of gradient. A gradient made with an micrographs. Convolutions or protrusions from giant 8-p (pore size) filtrate of granules was similar to the ly- granules, seen as small profiles, might explain those ob- sate gradient. servations. The authors ascribe the difference between their observation and those of others who have not Photomicrographs demonstrated such normal granules to technique.4'-2 Genetic variation within the Chediak-Higashi syn- Figure 4 is a photomicrograph of a Chediak-Higashi drome might account for these differences. No one has syndrome neutrophil and eosinophil. The staining reported a normal number of normal appearing shows characteristic giant granules in both, typical eo- peroxidase-positive granules in Chediak-Higashi syn- sinophil granules (Figure 4B), and no other cytoplas- drome neutrophils. Some studies of the fine structure mic granules in the neutrophil (Figure 4A). Character- of Chediak-Higashi syndrome granules are limited by istic azurophilic granules are absent from the the absence of peroxidase cytochemistrylll4 and in an- neutrophil, and the specific granules, which do not stain other there is the absence of detailed analysis of the with Romanowsky stains, are not seen. Neither purported normal azurophil granules."2 For example, azurophilic nor basophilic is a correct description of there was no quantitation of normal azurophil gran- the blue-grey or slate-grey color of the giant inclusion ules or analysis of the different types of azurophil gran- granules.34'35 Precedent favors the continued use of ules in normals for the purpose of comparison with the "azurophilic" to describe them, if not their color, be- rare small peroxidase-positive structures in some planes cause it has become synonymous with "peroxidase- which were called normal azurophil granules.12 These positive," which they are. The contrast between the color investigators also offer evidence that the massive inclu- 186 WEST AJP * January 1986 sions enlarge during development by engulfing small with rabbit heterophil specific granules,38 this associa- azurophilic granules, specific granules, and cytoplas- tion is not true for any granule of normal human neu- mic contents, and assert that this converts them to "sec- trophils.25'57'5 The present studies provide no support ondary lysosomes."''2 If this occurs, it might be consid- for a granule location for alkaline phosphatase. They ered aberrant phagolysosome formation, in which confirm a twofold and suggest a threefold higher ac- internal structures, rather than ingested particles, coa- tivity in these neutrophils, as compared with normal. 16 lesce with a lysosome. Localized to a broad density range approximately There is little hard evidence supporting any normal 1.14-1.15 g/ml, the activity marks a membrane fraction, azurophil granules in circulating neutrophils in the Che- as it does in normals.2' What membrane is increased diak-Higashi syndrome. Rausch et al did not find any in the Chediak-Higashi syndrome neutrophil is un- normal-appearing granules stainable with markers for known. azurophilic granules.47 The less dense human neutrophil Azurophilic and specific granules require careful azurophil granule normally contains elastase,4' and definition, especially when applied to Chediak-Higashi elastase has been demonstrated to be absent from Che- syndrome neutrophils. The light pink-lilac color of the diak-Higashi syndrome neutrophils. 18 Davis et al cytoplasm of normal neutrophils is not azure, but the showed no normal azurophilic granules in Ch- granules which give normal neutrophil cytoplasm its diak-Higashi syndrome mink.49"0 The present granule color are azurophilic, by convention. The granules of separation data provide no evidence for normal neutrophilic promyelocytes are truly azurophilic, hav- azurophil granules in Chediak-Higashi syndrome neu- ing a literal affinity for the azure-appearing dyes of the trophils. No peroxidase or other azurophil granule- Romanowsky method. However, virtually no such gran- associated enzyme activity sediments to the densities ule is known to exist in the mature neutrophil, although at which the two populations of azurophilic granules the term has been retained, accepted, and used as a syn- have been described.25.26'31'51 Therefore, the brothers onym for "peroxidase-positive."'31'39 Absent from cyto- whose blood we studied did not contain normal plasm of Chediak-Higashi syndrome neutrophils is the azurophil granules. Narrowly interpreted, these results usual light pink-lilac color which is associated with the could be said to be correct only for the two brothers presence of normal azurophilic granules; neither are who were available for study. Some clinical features sug- there present any actual azure-staining granules. These gest they are not typical of the Chediak-Higashi syn- cells have an empty look with the exception of the giant drome. In contrast to others, they are still alive at ages inclusion granules, which are stained neither azurophilic 35 and 34 years; they have been successfully treated for (in the actual or conventional sense) nor basophilic; recurrent pyogenic infections; and they have not devel- their color is intermediate, but is called azurophilic. oped the accelerated Chediak-Higashi syndrome, which Both by convention and because the immense cytoplas- usually results in childhood or adolescent death in those mic granules are peroxidase-positive, "azurophilic" re- diagnosed early who survive recurrent infections. These mains an appropriate adjective for "giant granule." brothers, about whom so much has been written as These observations are consistent with the absence of if it were applicable to all affected persons, are unique both normal azurophilic granules from density gra- and represent long-lived variants. This variation might dients of these patients' filtered neutrophil lysates. The explain the favorable response of another Che- absence of normal azurophilic granules from the cir- diak-Higashi syndrome patient to ascorbic acid,52 as culating neutrophils of these patients is probably im- compared with them.53 The implied genetic variation portant to their inability to respond normally to infec- within the Chediak-Higashi syndrome, however, is tions and malignancy.2.6.60 only clinical at present; it has no other scientific basis The giant azurophilic granule in neutrophils is com- in the present report. Nonetheless, reassessment of parable to the one in eosinophils. In eosinophils, how- data about the Chediak-Higashi syndrome which are ever, distinctive normal granules coexist with the giant largely2 51 l 1,15.16,24,34.35 or solely based upon them32"53-" granules. Neither giant inclusion is the same color as is needed. A condition in a child, comparable to the normal azurophilic granules in normal neutrophils. Chediak-Higashi syndrome because of large azuro- Until recently, lactoferrin had been localized uniquely philic neutrophil inclusions and recurrent infections, to specific granules57 61; using that premise, Rausch et was recently described and yet distinguished from it by a147 demonstrated by immunofluorescence that lactofer- several other abnormalities.'6 Characterization of var- rin is present in the Chediak-Higashi syndrome giant iation in the Chediak-Higashi syndrome and related azurophilic granules and concluded that they were not conditions has yet to be accomplished. correctly described as giant primary granules. However, Although alkaline phosphatase has been associated since Parmley et a162 showed that lactoferrin in normal Vol. 122 * No. 1 CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 187 neutrophils is in one of two types of primary or trophil. Important observations have been made in a azurophilic granule, it is no longer held that lactofer- few patients who lack "specific" granules.29 30'63 Now, rin is limited uniquely to the specific granule. Moreover, however, there appears to be no better example of dis- in contrast to normal granules, some gaint azurophilic ease marked by the absence of normal azurophilic gran- granules have structures consistent with specific granules than the Chediak-Higashi syndrome. ules in them."2 That evidence indicates that the giant azurophilic granules are not simply large primary granules and is consistent with our results. How specific granule profiles are incorporated into enlarging giant References azurophilic granules by a process which could be described as internally directed aberrant phagolysosome 1. Beguez-Cesar A: Neutropenia cronica maligna familiar formation and what the or con granulaciones atipicas de los leucocitos. Bol Soc precursor parent granule is Cubana Pediatr 1943, 15:900-922 for this process of lysosomal fusion are both unknown. 2. Blume RS, Wolff SM: The Chediak-Higashi syndrome: The results are preliminary regarding the gradient Studies in four patients and a review of the literature. density of the giant azurophilic granules; they appear Medicine 1972, 51:247-280 3. Windhorst DB, White JG, Zelickson AS, Clawson CC, to sediment to Q = 1.25-1.27 g/ml, a unique site in any Dent PB, Pollara B, Good, RA: The Chediak-Higashi comparable gradient, more dense than any previously anomaly and the Aleutian trait in mink: Homologous described granule, including eosinophil granules, the defects of lysosomal structure. Ann NY Acad Sci 1968, 155:818-846 most dense previously described.25'26 The wide range 4. White JG: The Chediak-Higashi syndrome: A possible of densities (1.25-1.27 g/ml) included in the visible band lysosomal disease. Blood 1966, 28:143-156 and the associated peroxidase and lysozyme activities 5. Clark RA, Kimball HR: Defective granulocyte chemotaxis in the Chediak-Higashi syndrome. J Clin Invest 1971, are consistent with the variation in size of the abnor- 50:2645-2652 mal granules. Filtration is more important than lysis 6. Root RK, Rosenthal AS, Balestra DJ: Abnormal bacteri- in destroying giant azurophilic granules. Their promi- cidal, metabolic, and lysosomal functions of Chediak- Higashi syndrome leukocytes. J Clin Invest 1972, 51: nence in monodispersed lysate suspensions shows that 649-665 they qualitatively survive lysis, although their suscep- 7. Stossel TP, Root RK, Vaughn M: Phagocytosis in chronic tibility to the shear forces of lysis has been observed. 16 granulomatous disease and the Chediak-Higashi syndrome. N Engl J Med 1972, 286:120-123 Although the granules are smaller than the 5-,u pores, 8. Clawson CC, White JG, Repine JE: The Chediak-Higashi enzyme and gradient data show destruction occurs pri- syndrome: Evidence that defective leukotaxis is primar- marily during passage through the 5-1A (pore size) filter. ily due to an impediment by giant granules. Am J Pathol 1978, 92:745-754 Perhaps their size is too close to that of the pores, which 9. Clawson CC, Repine JE, White JG: The Chediak-Higashi create destructive shear forces on them as they pass. syndrome: quantitation of a deficiency in maximal bac- The size of the giant azurophilic granule and its mem- tericidal capacity. Am J Pathol 1979, 94:539-548 10. Sadan N, Yaffe D, Rozenszajn L, Adar H, Soroker B, brane, similar or identical to plasma membranes known Efrati P: Cytochemical and genetic studies in four cases to be defective,21-24 are two possible "explanations." of Chediak-Higashi-Steinbrinck syndrome. Acta Hae- "Specific" granules are peroxidase-negative, comprise matol 1965, 34:20-29 11. Douglas SD, Blume RS, WolffSM: Fine structural studies band C sedimenting to Q = 1.18 g/ml, nonexclusively of leukocytes from patients and heterozygotes with the contain lysozyme, and have secretory functions.25-30 Chediak-Higashi syndrome. Blood 1969, 33:527-540 Chediak-Higashi syndrome-specific granules are nor- 12. White JG, Clawson CC: The Chediak-Higashi syndrome: The nature of the giant neutrophil granules and their in- mal by some assessments, including some interpreta- teractions with cytoplasm and foreign particulates. Am tions of electron micrographs"4 and the present reports J Pathol 1980, 98:151-196 of normal peak lysozyme specific activity in band C 13. Spicer SS, Sata A, Vincent R, Eguchi M, Poon KC: Lyso- some enlargement in the Chediak-Higashi syndrome. Fed and the absence of myeloperoxidase at band C (Q = Proc 1981, 40:1451-1455 1.18 g/ml). In contrast, the greater width and intensity 14. Fittschen C, Parmley RT, Austin RL, Crist MM: Vicinal of band C in the present report and other interpreta- glycol-staining identifies secondary granules in human normal and Chediak-Higashi neutrophils. Anat Rec 1983, tions of electron micrographs"4 suggest an increased 205:301-311 number of granules that are peroxidase-negative. With 15. Wolff SM, Dale DC, Clark RA, Root RK, Kimball HR: two distinct granule profiles having been previously de- The Chediak-Higashi syndrome: Studies of host defenses. Ann Intern Med 1972, 76:293-306 scribed at band C,25 biochemical markers to distinguish 16. Kimball HR, Ford GH, Wolff SM: Lysosomal enzymes more than one type of band C, "specific," peroxidase- in normal and Chediak-Higashi blood leukocytes. J Lab negative granule are needed to clarify normal neutrophil Clin Med 1975, 86:616-630 17. Zabucchi G, Cramer R, Soranzo MR, Tamaro P, Panizon contents as well as the possible increased number of F: Biochemical studies on the leukocytes in Chediak- specific granules in the Chediak-Higashi syndrome neu- Higashi syndrome. Acta Haematol 1977, 58:50-57 188 WEST AJP * January 1986

18. Vassalli J-D, Granelli-Piperno A, Griscelli C, Reich E: ing sugars on protein determination by the Lowry proce- Specific protease deficiency in polymorphonuclear leu- dure. Anal Biochem 1970, 34:591-593 kocytes of Chediak-Higashi Syndrome and beige mice. 38. Baggiolini M, Hirsch JG, de Duve C: Resolution of gran- J Exp Med 1978, 147:1285-1290 ules from rabbit heterophil leukocytes into distinct popu- 19. Goldfarb RH, Timonen T, Herberman RB: Production lations by zonal sedimentation. J Cell Biol 1969, of plasminogen activator by human natual killer cells: 40:529-541 Large granular lymphocytes. J Exp Med 1984, 159: 39. Litwack G: Photometric determination of lysozyme ac- 935-951 tivity. Proc Soc Exp Biol Med 1955, 89:401-403 20. Kanfer JN, Blume RS, Yankee RA, Wolff SM: Altera- 40. Talalay P, Fishman WH, Huggins C: Chromogenic sub- tion of sphingolipid metabolism in leukocytes from pa- strates: II. Phenolphthalein glucuronic acid as substrate tients with the Chediak-Higashi Syndrome. N Engl J Med for the assay of glucuronidase activity. J Biol Chem 1946, 1968, 279:410-413 166:757-772 21. Oliver JM, Zurier RB: Correction of characteristic ab- 41. Fishman WH: P-Glucuronidase, Methods of Enzymatic normalities of microtubule function and granule mor- Analysis. Edited by HU Bergmeyer. New York, Academic phology in Chediak-Higashi Syndrome with cholinergic Press, 1965, pp 869-874 agonists. J Clin Invest 1976, 57:1239-1247 42. The colorimetric determination of phosphatase, Techni- 22. Haak RA, Ingraham LM, Baehner RL, Boxer LA: Mem- cal Bulletin 104. Sigma Chemical Co., St. Louis, Mo, 1978, brane fluidity in human and mouse Chediak-Higashi leu- pp 1-25 kocytes. J Clin Invest 1979, 64:138-144 43. Appelmans F, Wattiaux R, de Duve C: Tissue fraction- 23. Ingraham LM, Burns CP, Boxer LA, Baehner RL, Haak ation studies: V. The association of acid phosphatase with RA: Fluidity properties and composition of erythrocyte a special class of cytoplasmic granules in rat liver. Bio- membranes in Chediak-Higashi Syndrome. J Cell Biol chem J 1955, 59:438-445 1981, 89:510-516 44. Chen PS, Jr, Toribara TY, Warner H: Microdetermina- 24. Nath J, Flavin M, Gallin JI: Tubulin tyrosinolation in tion of phosphorus. Anal Chem 1956, 28:1756-1758 human polymorphonuclear leukocytes: Studies in nor- 45. DeChatelet LR, Cooper MR: A modified procedure for mal subjects and in patients with the Chediak-Higashi the determination of leukocyte alkaline phosphatase. Bio- Syndrome. J Cell Biol 1982, 95:519-526 chem Med 1970, 4:61-68 25. West BC, Rosenthal AS, Gelb NA, Kimball HR: Separa- 46. West BC: Heparin inhibition of human neutrophil and tion and characterization of human neutrophil granules. eosinophil-enriched leukocyte acid P-glycerophosphatase. Am J Pathol 1974, 77:41-66 Proc Soc Exp Biol Med 1985, 178:373-384 26. West BC, Gelb NA, Rosenthal AS: Isolation and partial 47. Rausch PG, Pryzwansky KB, Spitznagel JK: Im- characterization of human eosinophil granules: Compar- munocytochemical identification of azurophilic and ison to neutrophils. Am J Pathol 1975, 81:575-588 specific granule markers in the giant granules of Chediak- 27. Wright DG, Bralove DA, Gallin JI: The differential Higashi neutrophils. N Engl J Med 1978, 298:693-698 mobilization of human neutrophil granules: Effects of 48. O'Connell RT, Anderson BR: Characterization of canine phorbol myristate acetate and ionophore A23187. Am neutrophil granules. Infect Immun 1982, 38:351-359 J Pathol 1977, 87:273-284 49. Davis WC, Spicer SS, Greene WB, Padgett GA: Ultra- 28. Wright DG, Gallin JI: Secretory responses of human neu- structure of bone marrow granulocytes in normal mink trophils: Exocytosis of specific (secondary) granules by and mink with the homolog of the Chediak-Higashi trait human neutrophils during adherence in vitro and dur- of humans: I. Origin of the abnormal granules present ing exudation in vivo. J Immunol 1979, 123:285-294 in the neutrophils of mink with the C-HS trait. Lab In- 29. Gallin JI, Fletcher MP, Seligmann BE, Hoffstein S, Cehrs vest 1971, 24:303-317 K, Mounessa N: Human neutrophil-specific granule 50. Davis WC, Spicer SS, Greene WB, Padgett GA: Ultra- deficiency: A model to assess the role of neutrophil- structure of cells in bone marrow and peripheral blood specific granules in the evolution of the inflammatory re- of normal mink and mink with the homologue of the sponse. Blood 1982, 59:1317-1329 Chediak-Higashi trait of humans: II. Cytoplasmic gran- 30. Gallin JI: Neutrophil specific granules: A fuse that ig- ules in eosinophils, basophils, mononuclear cells, and nites the inflammatory response. Clin Res 1984, 32: platelets. Am J Pathol 1971, 63:411-432 320-328 51. Kinkade JM, Pember SO, Barnes KC, Shapira R, Spitz- 31. West BC, Dunphy CH, Moore CA: Human neutrophil nagel JK, Martin LE: Differential distribution of distinct N-acetyl-p-D-glucosaminidase: Granule localization. Fur- forms of myeloperoxidase in different azurophilic gran- ther evidence for two azurophil granules. J Lab Clin Med ule subpopulations from human neutrophils. Biochem 1984, 104:60-68 Biophys Res Commun 1983, 114:296-303 32. Weary PE, Bender AS: Chediak-Higashi syndrome with 52. Boxer LA, Watananbe AM, Rister M, Besch HR, Allen severe cutaneous involvement. Arch Intern Med 1967, J, Baehner RL: Correction of leukocyte function in 119:381-386 Chediak-Higashi Syndrome by ascorbate. N Engl J Med 33. Boyum A: Isolation of mononuclear cells and granulo- 1976, 295:1041-1045 cytes from human blood. Scand J Clin Lab Invest 1968, 53. Gallin JI, Elin RJ, Hubert RT, Fauci AS, Kaliner MA, 21(Suppl 97):77-89 Wolff SM: Efficacy of ascorbic acid in Chediak-Higashi 34. Blume RS, Bennett JM, Yankee RA, Wolff SM: Defec- Syndrome (CHS): Studies in humans and mice. Blood tive granulocyte regulation in the Chediak-Higashi syn- 1979, 53:226-234 drome. N Engl J Med 1968, 279:1009-1015 54. Gallin JI, Klimerman JA, Padgett GA, WolffSM: Defec- 35. Blume RS, Padgett GA, Wolff SM, Bennett JM: Giant tive mononuclear leukocyte chemotaxis in the Chediak- neutrophil granules in the Chediak-Higashi Syndrome Higashi syndrome of humans, mink, cattle. Blood 1975, of man, mink, cattle, and mice. Can J Comp Med 1969, 45:863-870 33:271-274 55. Bell TG, Meyers KM, Prieur DJ, Fauci AS, Wolff SM, 36. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Pro- Padgett GA: Decreased nucleotide and serotonin stor- tein measurement with the Folin phenol reagent. J Biol age associated with defective function in Chediak-Higashi Chem 1951, 193:265-275 syndrome cattle and human platelets. Blood 1976, 37. Rosenthal HL, Sobieszczanska WA: Influence of reduc- 48:175-184 Vol. 122 * No. 1 CHEDIAK-HIGASHI SYNDROME NEUTROPHILS 189

56. Newburger PE, Robinson JM, Pryzwansky KB, Rosoff 61. Leffel MS, Spitznagel JK: Association of lactoferrin with PM, Greenberger JS, Tauber AI: Human neutrophil dys- lysozyme in granules of human polymorphonuclear leu- function with giant granules and defective activation of kocytes. Infect Immun 1972, 6:761-765 the respiratory burst. Blood 1983, 61:1247-1257 62. Parmley RT, Takagi M, Barton JC, Boxer LA, Austin RL: 57. Spitznagel JK, Daldorf FG, Leffell MS, Folds JD, Welsh Ultrastructural localization of lactoferrin and iron- IRH, Cooney MH, Martin LE: Character of azurophil binding protein in human neutrophils and rabbit hetero- and specific granules purified from human polymor- phils. Am J Pathol 1982, 109:343-358 phonuclear leukocytes. Lab Invest 1974, 30:774-785 63. Gallin JI: Neutrophil specific granule deficiency. Ann Rev 58. Bretz U, Baggiolini M: Biochemical and morphological Med 1985, 36:263-274 characterization of azurophil and specific granules of human neutrophilic polymorphonuclear leukocytes. J Cell Biol 1974, 63:251-269 59. Bainton DF, Ullyot JL, Farquhar MG: The development Acknowledgments of neutrophilic polymorphonuclear leukocytes in human bone marrow. J Exp Mqed 1971, 134:907-934 The author thanks Dr. Harry R. Kimball and Nancy A. 60. White JG: Virus-like particles in the peripheral blood cells Gelb for their contributions, many physicians for their care of two patients with Chediak-Higashi syndrome. Can- of these patients, and Tommie Lue Maddox for clerical as- cer 1966, 19:877-884 sistance.