Indian Pediatrics 2003; 40:822-833 

Background: An increased susceptibility to infections has been observed in some patients with phenylketonuria (PKU), which is not well known whether it is due to alterations of plasma essential amino acid concentrations or to some other factors. Objective: This study is designed to establish B cell and T cell functions in 44 children with classical PKU and tetrahydrobiopterin (BH4) deficiencies and the effects of too high plasma phenylalanine (PA) concentrations (16.53 to 30.54 mg/dL) on the same parameters. Design: B and T cell functions of 33 children with classical PKU (divided into two groups based on fasting mean plasma PA concentrations: Group-I = 20.9 ± 3.7 mg/dL, Group-II = 3.8 ± 1.02 mg/dL), and 11 children with BH4 deficiencies (Group III) were studied. The results were compared between the groups and referenced with previously reported values from healthy controls. Results: Delayed type skin hypersensitivity responses to purified protein derivative (PPD) in Group I and phytohaemagglutinin (PHA) in Group I, III were lower than the other groups and healthy controls. Plasma IgG and IgM concentrations of Group I was lower than the reference values. Although mean serum zinc and iron levels of all patients were lower than published values of healthy children, zinc and iron deficiencies in Group I, III were much more prominent as compared to Group II. Conclusion: The somewhat low plasma IgG concentrations in Group I may be related to the very high plasma PA levels, however the role of zinc deficiency as a causal factor can not be ruled out. BH4 metabolism defects do not appear to affect the same parameters. Impaired delayed skin hypersensitivity responses in Group I and III can be explained by severe serum zinc deficiency. In the light of this study, we conclude that in order to establish a causal relationship between PKU and immune functions, further studies need to be conducted after the correction of micro-nutrient status of such children.

Key words: Children, Immune functions, Phenylketonuria, Tetrahydrobiopterin deficiencies

Phenylketonuria (PKU) is an inborn error of metabolism resulting from either a deficiency of hepatic enzyme phenylalanine hydroxylase (PAH) which converts phenylalanine into tyrosine or its cofactor tetrahydrobiopterin (BH4) leading to elevated plasma phenylalanine (PA) and decreased tyrosine levels(1). An increased susceptibility to infections due to decreased immunoglobulin levels has earlier been reported in phenylketonuric patients(2). We noted recurrent upper and lower respiratory system infections such as severe common cold, pharyngitis, tonsillitis, croup, bronchitis, pneumonia and otitis media in some of our patients.

Alterations in plasma amino-acid patterns, plasma protein levels and trace element nutriture have been demonstrated in children with PKU on PA restricted diets while main-taining normal plasma PA concentration(3-5). Restricted essential amino acid intake, altered dietary fat intake and trace mineral status strongly influence humoral and cellular immune functions(6). Animal studies have demonstrated that administration of high amounts of PA partially suppresses their humoral immune functions(7). Thus, too high or too low plasma PA levels might alter plasma ratios of aromatic and/or branched chain amino acids and might be deleterious to other biochemical or physiological processes(8). Additionally, a relationship between the cellular immune function and BH4 metabolism has been speculated. Although some evidence suggesting the involvement of BH4 with interleukin-2 production in T cell proliferation is present, the physiological and clinical role of BH4 in the immune function is poorly under- stood(9-12).

Therefore, this study was conducted to examine B and T cell functions in children with classical PKU and BH4 deficiencies and the effect of high plasma PA concentrations on the same parameters.

Subjects and Methods

Selection Criteria

Children with classical PKU and BH4 metabolism disorders being followed up at Nutrition and Metabolism Unit, Ihsan Dogramaci Children’s Hospital, Ankara were recruited for the study. All individuals and their families were informed of the study and written informed consent was taken. Patients with malnutrition or active infection were excluded. The subjects were classified into three groups. Group I consisted of 21 classical phenylketonuric patients who missed screen-ing in the newborn period, were diagnosed very late, and therefore had very high plasma PA levels (20.9 ± 3.7 mg/dL). Group II had 12 age-matched classical phenylketonuric patients who were well-controlled as they were diagnosed during newborn period and had recommended plasma PA levels (3.8 ± 1.02 mg/dL). Eleven (4 female, 7 male) children with tetrahydro-biopterin deficiencies formed Group III.

Clinical and Laboratory Evaluation

Using a modified infection scoring system(13), a score of 10 was given for major infections (pneumonia, meningitis, septi-cemia, septic arthritis) requiring hospitaliza-tion and 5 for minor infections (otitis media, bronchitis, pleurisy, skin infections etc.). Positive history of recurrent infections was defined when the patient’s infection score exceeded 25 points (equivalent to two major and one minor or one major and three minor infections) over three years.

During routine follow-up (determining plasma PA levels and diet manipulations), 10 mL of whole blood sample was taken by venipuncture from each patient for complete blood count including differential cell counts, hemoglobin, hematocrit, Serum IgG, IgM, IgA, total protein, albumin, and trace element levels (iron, copper, zinc, selenium). Anti-poliovirus antibodies, lymphocyte subgroups, isohemagglutinins and plasma PA level were also measured. Complete blood count was measured by using a "Coulter counter STKS" hemocytometer. Lymphocytes were counted with a 100× objective Olympus Microscope from the routine blood smear and calculated by counting 200 cells. The serum total protein and albumin levels were measured on a Hitachi 9/7 autoanalyser.

For trace element estimation, samples of blood were placed in tubes washed with nitric acid and de-ionized water. Serum samples were separated by the centrifugation and stored at –20ºC till analysis. Serum selenium (Se) levels were determined by the spectro-fluorometric method described by Lalonde, et al.(14). Serum zinc and copper were measured by an atomic absorption spectro-photometer (model 1200, Varian Techtron). Serum iron was determined by using a Coleman Junior II (model 6/20) spectro-photometer( 15).

Immunological Parameters

Serum immunoglobulins were measured by using commercially prepared anti-sera to IgG, IgA, IgM and TurboxTM protein analyzer (Orion Diagnostica, Espoo, Finland)(16,17). Antibodies to polioviruses were measured by determining the ability of serum samples to neutralize the infectivity for cell cultures in vitro for each of the three types of poliovirus (Microneutralization procedure) at Refik Saydam Hlfzlss Slhha Institute of Hygiene Center, Cell Culture and National Polio Laboratory(18). A titer of 118 or above was considered positive.

Delayed type skin hypersensitivity was tested with purified protein derivative (PPD, Refik Saydam Hifzissihha Institution, 3 U/0.1 mL, Candida antigen (Hollister-Stier Labs 50 PNU/0.1 and PHA-P (phytoheamagglutinin P, Burroughs-Wellcomme Labs, 50 PNU/0.1 mL). An area of induration of 5 mm or more at 48 hours was considered a positive response. Lymphocytes subsets were analysed on a flowcytometer (Becton Dickinson) using Anti-CD3/CD4, Anti-CD3/CD8, Anti-CD3/CD19 and Anti-CD3/CD16+CD56 (FITC/PE) anti-body kits (Becton Dickinson, San Jose, CA)(19).

Statistical Analysis

Delayed skin hypersensitivity responses to PPD, PHA, and Candida of the groups were compared using Chi-square test. Age, weight, height, age at diagnosis, plasma PA level, serum trace element levels, immuno-globulins, cell counts of the groups were compared using Kruskal-Wallis one-way ANOVA. Whenever any statistically signifi-cant difference with respect to immune function parameters among three groups was found, the groups were compared with each other using Mann-Whitney U test to find out the groups creating the significant difference. Immune function parameters of children with PKU and BH4 metabolism disorders were also compared to published normal values for healthy children using Mann- Whitney U test. The individual relations between the positive history of recurrent infection, plasma PA level, serum trace elements level and immune functions were investigated by using Spearman’s rho correlation coefficients as the data distributions were not normal.

Results

The median ages in three groups were 96 months, 25 months, 48 months respectively. The groups were comparable with respect to age, height, weight and sex. The median age at diagnosis of PKU in Group I was 28 months as compared to 2 months and 6 months in Group II and III, respectively. As expected, median serum PA level of Group I (21 mg/dL 16.5-30.5) was significantly higher than the other two groups (Group II = 3.4 mg/dL; Group III = 3.6 mg/dL. There was no statisti-cal difference among the groups with respect to positive history for recurrent infections (Group I: 14.3%; Group II: 25%; Group III: 9.1%).

Comparison of Immunological Parameters amaong the Study Groups

Hematological data, serum immuno-globulin levels, lymphocyte subsets and the trace element levels are summarized in Table I. All these results were similar among children with different plasma PA concentrations and BH4 metabolism disorders. No significant inter group differences were found with respect to plasma levels of albumin, hemoglobin, hematocrit, IgG, IgM, IgA, total lymphocyte counts, percentages and number of CD3, CD4, CD8, CD19, CD16/56 and CD4 to CD8 ratio. Mean serum iron, copper and selenium levels were not significantly different among the groups. However, in well-controlled PKU children (Group II), mean serum zinc level was significantly higher than those of the other two groups.

Table I

Hematological, Serum Immunoglobulins Data, Total Number of Lymphocytes and 
T-Lymphocyte Subsets and Serum Trace Element levels in Classical PKU Children 
with Different Plasma PA Concentrations and in BH4 Metabolism Disorders*

	Classicial PKU				BH4 metabolism disorder
	Group I (n = 21)		Group II (n = 12)		Group III (n = 11)		P value
Hemoglobin (g/dL)	11.80	(9-13.7)	11.75	(10.4-13)	11.80	(9-16.3)	0.950
Hematocrit (%)	35.1	(27-41.2)	34.7	(30.2-38)	34.6	(29.3-57)	0.926
Albumin (g/dL)	5	(4.4-5.3)	5	(4.2-5.4)	5	(4.2-5.4)	0.397
IgG (mg/dL)	840	(450-2070)	955	(720-1620)	790	(490-1370)	0.490
IgM (mg/dL)	96	(40-313)	90	(41-163)	115	(63-412)	0.441
IgA (mg/dL)	127	(32-317)	112	(38-399)	93	(30-376)	0.406
Total lymphocytes (109/L)	4.94	(1.65-11.4)	5.6	(1.53-12.33)	6.5	(2.0-11.5)	0.379
CD32
% Lymphocyte	0.72	(0.56-0.79)	0.68	(0.54-0.78)	0.69	(0.53-0.76)	0.291
109/L	2.55	(1.3-4.2)	2.57	(0.78-3.34)	2.69	(1.77-5.12)	0.687
CD4
% Lymphocyte	0.37	(0.27-0.53)	0.39	(0.24-0.52)	0.36	(0.31-0.49)	0.993
109/L	1.39	(0.64-2.9)	1.44	(0.54-2.2)	1.48	(0.8-2.61)	0.758
CD8
% Lymphocyte	0.27	(0.19-0.49)	0.24	(0.15-0.41)	0.27	(0.18-0.36)	0.478
109/L	1.03	(0.49-1.65)	0.89	(0.22-1.77)	0.97	(0.87-2.41)	0.489
CD19
% Lymphocyte	0.15	(0.06-0.36)	0.19	(0.12-0.30)	0.19	(0.11-0.21)	0.317
109/L	0.50	(0.18-2.08)	0.83	(0.14-1.11)	0.71	(0.29-1.12)	0.480
CD16/56
% Lymphocyte	0.10	(0.02-0.24)	0.13	(0.06-0.16)	0.10	(0.05-0.25)	0.574
109/L	0.36	(0.09-0.66)	0.45	(0.07-0.83)	0.41	(0.22-0.85)	0.503
CD4/CD8 ratio	1.34	(0.55-2.68)	1.43	(0.68-2.94)	1.24	(0.9-2.69)	0.542
CD4%/CD8(%)
ratio	1.34	(0.54-2.54)	1.41	(0.59-2.78)	1.45	(0.94-2.68)	0.778
Iron (µg/dL)	42	(9-119)	68	(9-96)	18	(9-113)	0.194
Copper (µg/dL)	115	(40-238)	122	(42-323)	123	(58-173)	0.752
Zinc (µg/dL)	57	(30-134)	85	(60-100)	69	(27-92)	0.001
Selenium (µg/dL)	59	(41-117)	68	(48-101)	64	(54-80)	0.324

* Data are given as Median (Range)

 

Delayed skin hypersensitivity responses to purified protein derivative (PPD) in Group I were significantly lower (9.5%) than those of other two groups (83.3% and 63.6%) (Table II). Delayed skin hypersensitivity responses to PHA in Group I (23.8%) and III (27.3%) were significantly lower than Group II (83.3%). There was no significant inter group difference with respect to delayed hypersensitivity responses to Candida antigen. Group I revealed a noticeably low response to both PPD and PHA. In Group III, the response to PHA was found to be significantly low.

Table II

Positive Response to The Skin Tests and Positive Isohaemagglutinins and 
Anti-poliovirus Antibody Responses in The Study Population and published
 results for similar aged healthy children(25)

	Group I n = 21	Group II n = 12	Group III n = 11	Ref. (25) n = 13	X2  values
					I-II     I-III     II-III
PPD	2/21*	10/12	7/11	7/13	0.000	0.001	0.283
PHA	5/21*	10/12	3/11*	10/13	0.001	0.163	0.007
Candida	12/21	9/12	9/11	8/13	0.305	0.830	0.692
Isohaemagglutinins	19/19	12/12	11/11		-	-	-
Anti poliovirus antibody	20/21	12/12	11/11		-	-	-

*Significantly different from the reference (P<0.05). Ref.: published results from similar aged healthy
 children.
PHA (Gr. III - Ref.): p = 0.015; PHA (Gr.I - Ref.): p = 0.002; PPD (Gr.II - Ref.): p = 0.004.

 

The number of the cases having values below the normal limits in each study group were compared. All cases in Group II and III had normal (between +2 SD and –2 SD) serum IgG levels. However in Group I, 7 patients (33%) had serum IgG levels below the normal limit for the aged-matched healthy children(17) which was significantly different from other groups (P <0.05). Lymphocyte subsets in almost all patients in all groups were within the reference range(20). The number of the patients having subnormal serum zinc (Group I = 18; Group II = 3; Group III = 8) and iron (Group I = 13; Group II = 3; Group II = 9) levels(21) were also compared. In Groups I and III, significantly more patients had subnormal zinc and iron levels (P < 0.05) than Group II while there was no significant difference between Groups I and III (P = 0.371 for zinc and P = 0.248 for iron).

Comparisons of the Results of the Study Population to Published Results of

Similar-aged Healthy Children

Comparison of serum trace element levels, hematological and immunoglogical parameters between patients and published reference values from similar aged healthy children (17, 20-26) are shown in Table III. Children with PKU and BH4 metabolism disorders had lower IgG, zinc, iron levels and higher lymphocyte number. Although there was no significant difference between the study population and reference values, hemoglobin concentrations and hematocrit levels in the study population were found to be somewhat lower.

Table III

Hematological and Immunological Parameters of 44 children with Hyperphenylalaninemia
 Compared to Reference Values of Similar-Aged Childred (17,20-26)

	Study group (n = 44) Median (min-max)	Reference values Median (–2SD ± SD)	P values
Hematocrit	35 (27-57)	36 (32-44)21	0.078
Hemoglobin (g/dL)	11.8 (9.16.3)	12 (10.5-15.4)21	0.059
Total
Lymphocyte numb. (109/L)	5.6 (1.5-12.3)+	3.3 (1-11.9)20	0.001
CD3
109/L	2.6 (0.8-5.1)	2.4 (0.7-8)20	0.065
% Lymphocyte	0.69 (0.53-0.79)	0.66 (0.39-0.78)20	0.057
CD4
109/L	1.4 (0.5-2.9)	1.3 (0.3-5.5)20	0.611
% Lymphocyte	0.37 (0.24 - 0.53)	0.38 (0.23-0.58)20	0.674
CD8
109/L	1.0 (0.2-2.4)	0.9 (0.2-2.3)20	0.645
% Lymphocyte	0.26 (0.15 - 0.49)	0.25 (0.09-0.39)20	0.114
CD19
109/L	0.7 (0.14-2.1)	0.75 (0.1-3.1)20	0.190
% Lymphocyte	0.17 (0.06-0.37)	0.21 (0.08-0.44)20	0.246
CD16/56
109/L	0.37 (0.07-0.85)	0.30 (0.1-1.2)20	0.821
% Lymphocyte	0.10 (0.02-0.25)	0.10 (0.02-0.31)20	0.524
CD4/CD8 (%)	1.4 (0.5-2.8)	1.7 (0.9-3.8)20	0.056
Ig G (g/L)	8.3 (4.5-20.7)*	11.4 (5.4-17.1)17	0.037
Ig A (g/L)	1.1 (0.3-4)	1.1 (0.2-2.5)17	0.115
Ig M (g/L)	0.94 (0.4-4.1)	1.4 (0.6-1.98)17	0.409
Copper (µg/dL)	115 (40-329)	11322,23(75-145)21	0.336
Selenium (µg/L)	64 (40-117)	5524 (32.5-125.1)	0.131
Zinc (µg/L)	67.41 (27-134)*	10325(76-134)21	0.001
Iron (µg/L)	40 (9-119)	16026 (50-200)21	0.000

* Significantly lower than the reference values (P <0.05).
+ Significantly higher than the reference values (P <0.05).

 

Serum IgG and IgM levels of the Group I were significantly lower than the reference values while those of the Group II were similar to the reference values (Table IV). In addition, IgG levels of the Group III and IgM levels of the Group II were found to be somewhat lower than the reference values, although the difference was not statistically significant. Lymphocyte numbers were found to be higher in Group III when compared with reference values. Serum zinc and iron levels of all three groups were significantly lower than the reference values. Although there were no differences between the three groups, the average values of the hemoglobin and hematocrit of the Group I were found to be significantly lower than that of reference values (Table IV). Only three patients out of 21 had low hemoglobin level in the first group.

To view table IV please click the text  below Table IV

Delayed skin hypersensitivity responses to PPD and PHA in Group I, and PHA in Group III were found to be significantly lower than the results from our earlier report from similar aged healthy children(25).

Correlation between Individual Parameters

There was no statistically significant correlation between positive history of recurrent infection and other studied parameters. However, attenuated delayed skin hypersensitivity to PPD and PHA were found to be significantly correlated with the low serum zinc level (r = 0.456 and 0.508 respectively; P = 0.002) and the high plasma PA concentrations (r = –0.616 and –0.328 respectively, P = 0.000 and 0.030). There was also a weak negative correlation between plasma PA concentrations and serum zinc levels (r = –362, P = 0.016). In other words, the skin responses to PPD and PHA skin tests in patients who had zinc deficiency and high plasma PA concentrations were significantly lower than the responses of patients with normal serum zinc level and better plasma PA concentrations.

Discussion

The main goal of PKU treatment is to maintain plasma PA levels in the recommended range. Elevated plasma PA levels above the recommended level leads to retarded mental and motor development of the patient. The severity of impairment correlates with the maximum level of the PA and its duration. Infections in patients during treatment can increase the plasma PA level. Patients with PKU have been observed to be more susceptible to infections(2). As the infections can affect the response to the treatment of PKU, a critical question to address is whether these infections are caused by primary immune dysfunction.

Studies on the PKU and increased susceptibility to infections relationship do not suggest an increased susceptibility(7,8). Only some changes in humoral immunity have been shown. However, the data is scarce and include a limited number of patients. Imbalance of the essential amino acids can affect humoral immune functions(27). In experimental animals, it has been shown that L-phenylalanine decreased both antibody production and homograft rejection(7). However, a direct influence of elevated plasma PA levels on immune functions has not been reported. In our study, there was no significant difference between the groups except for plasma IgG concentrations and their response to the skin tests. In addition, among the studied parameters only serum zinc levels revealed a significant difference among the three groups. Interestingly, both serum zinc and iron levels in all 3 groups were found to be significantly lower than that of similar aged healthy children studied previously by the same method in the same laboratory. Lower serum zinc levels in the patients with PKU have earlier been reported(4,5). In our study, children with very high PA level (Group I) and BH4 metabolism disorders (Group III) had much lower serum zinc levels than well-controlled PKU children (Group II). This could be explained by poor intake of animal protein and zinc rich foods especially in Group I because of their relatively low socioeconomic status. Amount of total and animal protein in diet has been shown to have an effect on zinc absorption(28). Similarly, patients with BH4 metabolism disorders do not consume PA restricted artificial formula as much as patients in Group II. In another study, Taylor, et al.(29) found that serum zinc levels were subnormal in 42% of 19 patients with PKU and attributed the zinc deficiency to a low protein diet. In our study, although the patients in Group II were on higher amounts of artificial formulas relative to other groups, they had better serum zinc and other trace element levels than the other two groups. Since all patients with PKU have zinc deficiency, they should be tested for zinc levels and supplemented with zinc regularly.

Regarding iron deficiency, there was no difference in hemoglobin, hematocrit or serum iron between the study groups. When compared to normal standards, the mean Hb and hematocrit was somewhat lower in Group I (P = 0.043, P = 0.049) but was still within normal range. However, serum irons in all groups were significantly less than the standards. Thus, all patients with PKU are at risk of subclinical iron deficiency. The results of the serum copper and selenium levels in patients with PKU were similar to those of the age-matched healthy children. Hence, patients with PKU are not at risk for developing copper and selenium deficiencies while on diet.

There was a positive correlation between serum zinc level and the delayed skin hypersensitivity responses to PPD and PHA in our study. On the other hand, a negative correlation was found between plasma PA concentration and the delayed skin hyper-sensitivity responses to PPD or PHA. Group I revealed a noticeably lower response to both PPD and PHA. In Group III, the response to PHA was found to be significantly lower than the control group. Therefore the reason for the deterioration of the responses to the skin tests in these two groups could be related to the significantly low serum zinc level in two groups or to the combination of very high plasma PA levels and low serum zinc concentrations.

Serum IgG and IgM levels in Group I was found to be somewhat lower than reference values. But, the decrease in the IgG level in patients was not severe and is unlikely to be of clinical significance because there was no correlation between positive history of recurrent infections and immune functions. Serum zinc deficiency also can decrease serum immunoglobulin levels(30). However, in our study we did not find any correlation between serum zinc levels or PA concentrations and plasma IgG concentrations or other immune parameters. Thus, mild decrease in IgG levels observed in these patients might be related to high plasma PA levels.

There was no abnormality in immune functions in Group II who regularly used up much more synthetic formulas than the other two groups. Especially quantitative IgG levels were not lower than the reference values. Other important results were the high levels of trace elements and better delayed skin hypersensitivity responses to PPD and PHA in Group II than those of other two groups. In earlier studies in the literature, serum Ig levels in patients on diet treatment with formulas have been found to be significantly lower than the control group. It was suggested that low serum IgG levels may be due to the decreased levels of serum protein, amino acid, iron, zinc and selenium in patients fed by artificial formulas(2,8). Unlike the findings of previous studies, serum IgG levels and trace elements balance in our study were better in patients fed regularly with synthetic formulas than the patients fed without artificial formulas. This could be explained by changes made in the ingredients of the formulae by technical developments (personal communications with manufacturers) based on the results of previous studies. Under the lights of several recent published studies(31,32) especially in our country, it can be suggested that the risk of trace element deficiency is higher in PKU patients not receiving artificial formulas. This means that artificial formulas without PA or with restricted amount of PA have a somewhat positive effect on serum trace elements levels of patients with PKU.

In summary, we conclude that somewhat lower plasma IgG and IgM concentrations in Group I and impaired delayed skin hyper-sensitivity responses in Group I and III may be related to very high plasma PA levels; however the role of zinc deficiency can not be excluded. Therefore, in future studies it would be reasonable to first normalize serum trace elements by giving these patients diet supplementation prior to assaying immune functions. Then, the possible role of plasma PA levels on immune functions should be carefully investigated.

Acknowledgements

We thank to Murat Tuncer for his help on the assessment of lymphocyte subsets and to Serap Kalkanoglu for her help on taking blood samples from the patients and to all the technical staff involved in the study.

Contributors: TC conceived and designed the study, interpreted the data, TK coordinated the study, analyzed and interpreted the data, and wrote the manuscript, EQ did the assessment of antibodies to polioviruses, IO and FE interpereted the data, helped in drafting the paper.

Funding: This work was supported by Hacettepe University Research Fund.

Competing interest: None declared.