Pituitary

, Volume 16, Issue 1, pp 83–90

Involvement of genes related to inflammation and cell cycle in Idiopathic Short Stature

Authors

    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
    • Laboratory of Molecular and Cellular Endocrinology, Department of Internal MedicineUniversity of Turin
  • Flavia Prodam
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Giulia Genoni
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Francesca De Rienzo
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Gillian E. Walker
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Stefania Moia
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Stefania Riccomagno
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Simonetta Bellone
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
  • Gianni Bona
    • Division of Pediatrics, Department of Medical SciencesUniversity of Piemonte Orientale
Article

DOI: 10.1007/s11102-012-0378-8

Cite this article as:
Trovato, L., Prodam, F., Genoni, G. et al. Pituitary (2013) 16: 83. doi:10.1007/s11102-012-0378-8

Abstract

Idiopathic Short Stature (ISS) defines a condition in which height is <−2SD compared to the mean of a reference population where systemic, endocrinological, nutritional or chromosomal disorders have not been identified and diagnosis is based on exclusion of any known causes of short stature. JAK/STAT pathway is triggered by GH binding to the GH receptor and promotes cellular growth through transcription of GH-responsive genes. In order to identify “candidate genes” differently expressed in ISS subjects with respect to control ones, we analyzed the expression of 84 genes related to JAK/STAT pathway by RT2 Profiler PCR array approach in a total of 10 subjects. Then, we validated the observed data by Real Time PCR and ELISA assays in a major number of subjects. We found two genes that were differently expressed in ISS subjects with respect to the control group: CXCL9 and FCGR1A/CD64, both significantly up-regulated (fold change 2.17 and 1.70, respectively) and belonging to family of IFN-γ-inducible factors. Further, ISS subjects showed an increased gene expression of IFN-γ and IFI16, higher serum levels of IFN-γ but similar levels of CXCL9 when compared to healthy subjects. In addition, we showed a pubertal modulation of CXCL9 levels. These data suggest that inflammatory and regulatory factors of the cell cycle may be involved in the ISS condition, introducing a new perspective to its etiology.

Keywords

CytokinesChemokinesGene profileIdiopathic short statureCellular cycle

Introduction

Height is a complex trait regulated by multiple genetic and environmental factors. Idiopathic Short Stature (ISS) describes a heterogeneous group of children consisting of many presently unidentified causes of short stature; it is estimated that about 60–80% of all short children at or below −2 SDS fits the definition of ISS [1]. In addition, ISS children show variable IGF-I levels which are either normal or low, suggestive of peripheral GH resistance due to post-receptor defects in signal transduction, but not due to GH receptor (GHR) mutations [2]. It is known that only 5% of subjects with ISS present GHR mutations, moreover several studies in ISS children have failed to demonstrate mutations in either GH or GHR genes [3]. In order to understand the pathogenesis of ISS, other genes whose function is not associated to GH/GHR have been investigated, including PTPN11 and SHOX (Short-stature Homeobox gene on the X chromosome). Mutations in PTPN11 gene are commonly associated to the pathogenesis of Noonan Syndrome but are not a common cause of ISS [4], as also SHOX mutations in those ISS patients with subtle radiographic abnormalities [5]. Further, autoimmune mechanisms have been proposed as a possible cause of ISS, as suggested by the presence of Anti-Pituitary-Antibodies directed against GH-secreting cells in the serum of subjects with GH Deficiency or ISS [6].

GH signalling starts with the binding of GH to GHR that induces the JAK/STAT cascade promoting cellular growth and proliferation through the transcription of GH-responsive genes. The Janus Kinase (JAK)–Signal Transducer and Activator of Transcription (STAT) signal transduction pathway is activated by a large number of cytokines and growth factors, including interferon (IFN)-α/β/γ and interleukins, in addition to GH [7]. In mammals, the JAK/STAT pathway plays a significant role in mediating biological activities such as growth, survival, differentiation and resistance to pathogens [8]. Among the JAK family proteins, JAK2 is a key molecule activated by GH and several studies showed the importance of the association GHR/JAK2 for the downstream signalling [9]. Regarding the STAT family of proteins, GH induces specifically STAT1, STAT3, STAT5a and 5b but only STAT5b is involved in IGF-I activation [10]. In a previous investigation, STAT5b was assessed as a candidate gene for ISS, however, direct sequencing of the gene failed to show any alterations [11].

Recently, Genome Wide Association Studies (GWAS) showed that some genes not strictly related to the GH/IGF-I axis but implicated in other biological pathways, are linked to growth, such as the Hedgehog signaling pathway, bone morphogenic proteins (BMPs), genes encoding Extracellular Matrix Components (ECM) and many others [12].

Therefore, the aim of this study was to analyze the quantitative expression of 84 genes-related to the JAK/STAT pathway in ISS subjects with normal GH and IGF-I levels with respect to subjects with normal stature, in order to identify “candidate” genes that could address new fields of investigation. We used the RT2 Profiler PCR Array, a technology that tracks the expression of a profile of genes related to a biological or pathological processes. In particular, we focused on the JAK/STAT pathway, as it represents the most important signal transduction pathway activated by GH and then seemed more appropriate to investigate ISS condition.

Subjects and methods

Subjects

From May 2008 to May 2009, 41 consecutive healthy children and 60 ISS pediatric subjects admitted at the Division of Pediatrics, University of Piemonte Orientale, Novara, Italy, were enrolled in the present study. Informed parental consent and approval by the Local Ethical Committee were obtained before beginning the study.

Inclusion criteria were the presence of ISS defined as a height <−2SD with respect to the average population, a growth velocity below the 25th percentile for chronological age and a chronological or 6–18 months delayed bone age with a normal height and weight at birth and a normal GH response to provocative tests. Exclusion criteria were syndromes or chromosomal defects, skeletal dysplasia, severe neurological impairments, autoimmune diseases, hypopituitarism, in particular GH deficiency, past or present rhGH therapy, type 1 and 2 diabetes, renal dysfunction, alcohol or smoking and any other condition known to influence height and body composition (insulin, glucocorticoids, endocrine diseases like hypothyroidism or Cushing’s syndrome). Healthy children of normal height, matched for sex, age and pubertal stage but not tested for dynamic GH secretion, were used as controls.

Subjects underwent a complete auxological examination and auxological parameters were evaluated according to Italian growth charts [13]. Height and weight were respectively measured by using the Harpenden stadiometer and an electronic scale. Bone age was evaluated using the Tanner-Whitehouse method 2 (TW2). Body Mass Index (BMI) was calculated as body weight divided by squared height (kg/m2). BMI standard deviation score was assessed with LMS method [13]. Growth velocity was calculated per year (cm/year) with two consecutive measurements performed at least 6 months of follow-up. Birth length and weight were compared with neonatal standards [14]. All the ISS subjects performed a GH stimulation test which was defined as normal with a peak GH response above or equal 20 ng/ml after GHRH + arginine.

RT2 profiler PCR array

In this study a RT2 Profiler™ PCR array system “Human JAK/STAT Signaling Pathway” (SABioscience Corporation, USA) was used. The RT2 Profiler PCR Array was performed in five subjects with ISS and five healthy age-matched controls with normal stature who met the inclusion criteria and have been enrolled consecutively.

Auxological characteristics of ISS and control subjects evaluated for PCR array are summarized in Table 1. All subjects have a harmonic short stature with weight adequate to stature. The subject with BMI just under 3rd percentile have height on the 3rd percentile. None of them presented syndromes or chromosomal defects, skeletal dysplasia or chronic diseases.
Table 1

Auxological and hormonal characteristics of five ISS subjects analyzed by RT2 Profiler™ PCR Array

Gender

F

M

F

F

F

Age (years)

10.3

13.4

10.1

10.7

13.3

Tanner stage

1

2

1

2

3

Height (cm)

126.0

134.4

125.9

128.3

137.2

Height (SDS)

−2.1

−3.1

−2.2

−2.0

−3.5

Target Height (cm)

161,9

166,8

154,4

161.1

161.1

Target Height (percentile)

48°

12°

10°

43°

43°

Rate of growth (cm/year)

4.4

3.3

3.8

Rate of growth (percentile)

< 1°

12°

Bone age

No delay

< 3 years

No delay

No delay

< 1 year

BMI (kg/m2)

18.9

16.8

18.8

14.0

17.3

BMI (percentile)

60°

10°

60°

<3°

10°

GH stimulation test

GHRH + Arg

GHRH + Arg

GHRH + Arg

GHRH + Arg

GHRH + Arg

GH peak (ng/ml)

82.7

91.6

24.3

60.5

89.9

IGF-I basal (ng/ml)

241.7

152.0

120.8

208.4

94.0

IGF-I basal (SDS)

−0.007

−0.87

−0.55

−0.16

−0.95

The GH peak was studied in response to the GHRH + Arginine (GHRH + Arg) provocative test

BMI body mass index, F female, M male, SDS standard deviation score

Lymphocytes were separated from the whole blood by density gradient centrifugation and total RNA was extracted using a RT2 qPCR-Grade RNA Isolation Kit and retro-transcribed by RT2 First Strand Kit (SABioscience Corporation). The run was performed using the ABI PRISM 7000 Sequence Detection System under the following conditions setup provided by supplier. The data were normalized to the values for housekeeping gene panel using a ΔΔCt method, according to the manufacturer’s instructions (http://www.sabiosciences.com/pcrarraydataanalysis.php). Average Ct of all samples were above 25 and all comments were OKAY on the system control, then, they excluded a potential bias.

According to the manufacturer’s instructions that state that the requested number to obtain a significant result is at least three arrays for each evaluated group, five subjects for each group were considered sufficient for the screening with a RT2 Profiler™ PCR approach. Furthermore, the use of a SYBR green detection during the amplification makes this method more reliable than a microarray analysis allowing to reduce the sample size [15]. According to this, our group recently published a paper where the same approach was used [16].

Quantitative real time PCR (qRT-PCR)

In order to validate the results of the RT2 Profiler PCR Array, a qRT-PCR was performed in 20 ISS subjects: 5 children already tested with the PCR Array and 15 ISS additional subjects. Seven different control subjects were used as calibrators in following step of the analysis. Total RNA was extracted as described above and 2 μg of RNA was retro-transcribed using the High Capacity Reverse Transcription kit. The following TaqMan Gene Expression Assays were used: human CXCL9 (Hs00171065_m1), human FCGR1A/CD64 (Hs00174081_m1), human IFN-γ (Hs99999041_m1) and human IFI16 (Hs00194261_m1). Gene expression was normalized to human β-actin (Hs99999903_m1) and relative gene quantity was calculated using the comparative Ct method (2−ΔΔCt). Results are expressed as mean ± SEM of fold regulation with respect to all controls.

Hormonal and cytokine assays

GH during provocative tests and basal IGF-I were measured in the entire group using commercial RIA kits (SM-C-RIA-CT, Pantec, Italy). Serum levels of CXCL9 (range standard curve 0–2,000 pg/ml) and IFN-γ (range standard curve 0–1,000 pg/ml) were measured in 60 ISS and 41 healthy normal stature subjects, using Human CXCL9/MIG Quantikine ELISA kit and Human IFN-γ Quantikine ELISA kit, respectively (R&D Systems, Space Import Export, Milan, Italy).

Statistical analysis

Statistical analysis for RT2 Profiler PCR array was performed using Web-Based PCR Array Data Analysis and the significance was established at p ≤ 0.05. For the results of qRT-PCR, all data are expressed as means ± SEM or confidence interval (CI). For continuous variables, non-Gaussian data were log transformed prior to the analysis. The variation between groups was compared by means on Student’s, nonparametric Wilcoxon and Mann–Whitney U tests or ANCOVA, where appropriate. The stepwise regression model with two-tailed probability values and 95% confidence intervals (CI) was used to measure the strength of the association between variables. Statistical significance was assumed for at least p < 0.05. Statistical analyses were performed with SPSS for Windows version 17.0 (SPSS INC; Chicago, IL, USA).

Results

Profiling of JAK/STAT pathway

The expression of 84 genes related to the JAK/STAT pathway was compared between subjects with ISS or normal stature (control group) in order to identify candidate genes potentially linked to this condition. Serum levels of IGF-1 and stimulated secretion of GH were within normal ranges in all the ISS subjects (Table 1). Figure 1 shows respectively in panel a the “heat map” of the level of expression of 84 genes related to JAK/STAT pathway in the ISS subjects when compared to controls, while in panel b the values of fold regulation of all genes. We observed that in ISS subjects two genes are differently expressed: CXCL9, a chemokine ligand 9 (p = 0.037) and FCGR1A, Fc fragment of IgG, high affinity Ia, receptor (CD64; p = 0.019), both up-regulated with a fold regulation of 2.17 (95% CI 1.03–3.31) and 1.70 (95% CI 1.19–2.20), respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs11102-012-0378-8/MediaObjects/11102_2012_378_Fig1_HTML.gif
Fig. 1

Gene profiling of the JAK/STAT pathway. The expression of the 84 genes related to the JAK/STAT pathway was analyzed employing a qRT-PCR based RT2 Profiler™ PCR Array in five subjects with ISS and five subjects with normal stature. The gene expression was normalized to the average expression of 5 housekeeping genes (B2 M, HPRT1, RPL13A, GAPDH, ACTB) and directly calculated using the Web-Based PCR Array Data Analysis as indicate in Subjects and Methods paragraph. a The “heat map” represents the level of gene expression of all 84 genes related JAK/STAT pathway and was obtained by the mean of gene expression of 5 ISS and 5 control subjects; the green and red wells indicate the down and up-regulated genes, respectively, while the black wells indicate a similar gene expression between ISS and the control group (fold regulation = 1). b The table indicates the name and the fold regulation of all 84 genes, in red and in bold are indicated the genes whose expression was statistically significant (p ≤ 0.05) with respect to the control group

With respects to genes involved in the regulation of stature, such as GHR and JAK2, no significant variations were observed between ISS subjects and the control group (fold change 1.07, 95% CI 0.00001–2.44, p = 0.519; and 1.09, 95% CI 0.77–1.41, p = 0.579, respectively).

Gene expression analysis of CXCL9, FCGR1A/CD64, IFN-γ and IFI16

On the basis of the array results, the investigation was focused on the two up-regulated genes, CXCL9 and FCGR1A/CD64 with their up-regulation confirmed by qRT-PCR in a group of further 15 children with ISS (fold regulation 2.30 ± 0.17; p < 0.0001 and 1.50 ± 0.08; p < 0.0001, respectively) as shown in Fig. 2a, b.
https://static-content.springer.com/image/art%3A10.1007%2Fs11102-012-0378-8/MediaObjects/11102_2012_378_Fig2_HTML.gif
Fig. 2

Gene expression levels of CXCL9, FCGR1A/CD64, IFN-γ and IFI16. In order to validate array’s results, the qRT-PCR was performed in the same subjects evaluated for the gene expression profiling and additional 15 ISS subjects, both prepubertal and pubertal for a total of 20 subjects. a, b The gene expression of CXCL9 and FCGR1A/CD64, while c, d the gene expression of IFN-γ and IFI16. In all panels, the control corresponds to a subjects with a normal stature selected among those enrolled in the study and used as calibrator (n = 7). All 20 ISS subjects were analyzed respect to all 7 calibrator and the results are expressed as the mean ± SEM of fold regulation obtained by each single analysis and calculated by ΔΔCt method. Statistical analysis was performed using Student’s t test and significance was established for ***p < 0.0001

Because CXCL9 is an IFN-γ-induced chemokine and IFN-γ also stimulates FCGR1A/CD64, the involvement of this cytokine as a trigger potentially responsible for the higher expression of both molecules was hypothesized. By PCR array, we previously observed that INF-γ had a fold regulation of 1.84, but this result was not statistically significant. Therefore, to assess our hypothesis, we evaluated the expression of IFN-γ mRNA by qRT-PCR in the same group of subjects evaluated for CXCL9 and FCGR1A/CD64 and we found a higher expression of IFN-γ (fold induction 10.00 ± 0.18; p < 0.001) when compared with the control group, as represented in Fig 2c. The evidence of a higher IFN-γ expression of in ISS subjects, suggested that it could play an important function by exerting a direct action on genes related to it or specific cellular processes. Thus, hypothesizing an effect of IFN-γ on cell cycle, we evaluated IFI16, a gene IFN-γ-inducible involved in the regulation of cell cycle with a higher mRNA expression observed in ISS subjects with respect to controls (fold induction 1.50 ± 0.15; p < 0.0001; Fig. 2d).

Serum levels of IFN-γ and CXCL9

In order to explore the role of IFN-γ in ISS condition, we evaluated its serum levels in 60 ISS and 41 healthy subjects with normal stature. Auxological and biochemical characteristics are summarized in Table 2.
Table 2

Clinical and biochemical characteristics of control and ISS subjects evaluated for serum levels of INF-γ and CXCL9

 

CTRL

ISS

Variables

All

Prepubertal

Pubertal

All

Prepubertal

Pubertal

Subjects

41

16

25

60

45

15

Age (yrs)

10.1 ± 0.5*

7.6 ± 0.3

11.7 ± 0.7c

9.1 ± 0.4*

7.2 ± 0.4

11.6 ± 0.3c

Tanner stage

 

I (16)

II(13); III(3); V(9)

 

I (45)

II(9); III(1); V(5)

HSDS

0.5 ± 0.1

0.5 ± 0.2

0.5 ± 0.2

−1.9 ± 0.1***

−1.9 ± 0.1***

−1.8 ± 0.2***

BMISDS

−0.37 ± 0.19

−0.66 ± 0.35

−0.18 ± 0.22

−1.04 ± 0.13**

−1.10 ± 0.14

−0.70 ± 0.36*

IGF-I (ng/ml)

264.7 ± 30.0

213.0 ± 27.8

307.9 ± 44.4a

128.1 ± 8.9***

119.1 ± 9.7**

172.7 ± 12.9**,b

GH peak (μg/dl)

72.2 ± 5.0

71.7 ± 5.5

73.8 ± 12.4

IFN-γ (pg/ml)

3.7 ± 0.8

2.5 ± 1.0

4.5 ± 1.3

9.1 ± 1.0***

9.2 ± 1.1***

8.9 ± 2.7**

CXCL9 (pg/ml)

115.3 ± 21.8

151.0 ± 39.9

92.4 ± 24.7a

99.2 ± 15.2

113.1 ± 17.6*

30.1 ± 6.3*,b

Data are expressed as mean ± SEM

* Significantly different within subjects according to groups and subgroups (* p < 0.05; ** p < 0.01; *** p < 0.001); significantly different between prepubertal and pubertal of the same group (a p < 0.05; bp < 0.01; cp < 0.0001);  response to arginine + GHRH test

BMISDS body mass index standard deviation score, HSDS height standard deviation score, CTRL control group, ISS idiopathic short stature

Both prepubertal and pubertal ISS subjects showed higher serum levels of IFN-γ compared to healthy subjects with normal stature (9.1 ± 1.0 vs. 3.7 ± 0.8 pg/ml; p < 0.001). The significance was maintained when groups were weighted for confounding factors as age, gender or puberty. IFN-γ levels were similar with respect to the pubertal status (Table 2). INF-γ was found to be independently associated with ISS (beta: 0.470; CI 0.253–0.287) and ISS could explain the 47.2% of the variation of IFN-γ levels (p < 0.0001). Furthermore, INF-γ levels remained significantly associated with ISS also when adjusted for gender, age, BMISDS and IGF-I levels.

In addition, we also evaluated the circulating levels of CXCL9 that were similar in both groups (99.2 ± 15.2 vs. 115.3 ± 21.8 pg/ml; p > 0.05). However, evaluating subgroups, both prepubertal (113.1 ± 17.6 vs. 151.0 ± 39.9 pg/ml; p < 0.04) and pubertal ISS subjects (30.1 ± 6.3 vs. 92.4 ± 24.7 pg/ml; p < 0.01) had lower CXCL9 levels than observed in healthy subjects. CXCL9 levels were higher in prepubertal than pubertal individuals in both groups (p < 0.05).

CXCL9 levels were associated with pubertal status in the whole group (r: 0.200; p < 0.05), with increased strength when corrected for the ISS diagnosis (r: −0.234; p < 0.03).

IFN-γ and CXCL9 levels failed to correlate with each other in all groups and subgroups.

Discussion

According to previous findings, our study shows a minor participation of GHR alterations in the pathogenesis of ISS [17]. In addition, this study showed no significant variations in the gene expression of STATs and JAK2 between ISS and normal subjects. Moreover, mutations of STAT5b were reported in patients with GH insensitivity and primary IGF deficiency [18] and mutations of JAK2 were reported only in four cohorts of patients with myeloproliferative disorders [19]. On such basis, alternative mechanisms in the ISS condition should be explored.

In this exploratory pilot study, the analysis of 84 genes related to the JAK/STAT pathway revealed that in ISS subjects CXCL9 and FCGR1A/CD64 are significantly up-regulated with respect to subjects with normal stature. In addition, we also observed that IFN-γ had significantly higher mRNA and circulating protein levels. The evidence of higher IFI16 gene expression, a molecule participating in the suppression of cell cycle progression and cell survival, could potentially drive future researches on cell cycle disruption in ISS.

As previously reported, many evidences show that the relationship between the immune and endocrine system represents a bidirectional process [20]. In particular, the role of CXCL9 in the immune response has been well described. In fact, it is a key member for IFN-γ-inducible chemokines, along with CXCL10, CXCL11 and their receptor CXCR3, in the early stage of some autoimmune diseases that affect endocrine glands, in particular Hashimoto autoimmune thyroiditis and type 1 diabetes [21]. To date, there is no evidence regarding the role of CXCL9 in ISS or any other form of short stature, although, growth retardation is commonly seen in children with chronic inflammatory conditions such as juvenile idiopathic arthritis [22].

In this study, we also found that ISS subjects presented a higher gene expression of FCGR1A/CD64. Accordingly, an up-regulation of FCGR1A/CD64 expression has been observed in many inflammatory conditions, such as inflammatory bowel diseases [23]; a recent report includes it as a new candidate biomarker for pediatric infections [24].

Another interesting result of this study was the up-regulation of the IFN-γ gene expression. This cytokine has a dual role as pro- or anti-inflammatory cytokine and it is known that IFNs are potent inhibitors of cell growth both in vivo and in vitro and exhibit an antitumor activity [25]. Based on our results, we proposed that IFN-γ is responsible for the higher expression both of CXCL9 and FCGR1A/CD64 in ISS subjects, even though we are unable to explain why IFN-γ is increased. Moreover, we recorded not only an up-regulation of the IFN-γ gene but also higher circulating levels of this cytokine in ISS phenotype with respect to subjects with normal stature. Furthermore, IFN-γ levels were shown to be a strong predictor of ISS by correlation analysis.

The central role of IFN-γ proposed in this study could imply its involvement in cellular growth and in the cell cycle, then, based on this assumption we evaluated IFI16.

IFN-γ-inducible IFI16 is a human member of the IFN-inducible structurally related to the p200 family of protein which has been demonstrated to participate in the suppression of cell cycle progression and cell survival, as that mediates the growth-inhibitory activities of interferons. Furthermore, it has been implicated in transcriptional regulation by modulating protein–protein interactions with p53 tumor suppressor protein and other transcription factors [26]. Moreover, an increased IFI16 expression is associated with cellular senescence in human fibroblasts [27] and it’s the up-regulation in this process is mediated by p53 [28]. Our results showed that in ISS subjects both the IFI16 and INF-γ gene expression were increased compared to control subjects. It could therefore be hypothesized the existence of a disruption of cell cycle regulation in ISS, taking into account that our patients had a nearly to significance increase in p53 gene expression (data not shown). In vitro studies evaluating the cell cycle will be able to clarify if disruption of cell cycle progression can be related to the ISS phenotype.

In addition to the evaluation of IFN-γ serum levels described above, we also measured the serum levels of CXCL9 but our data suggested a lack of association between serum levels of IFN-γ and CXCL9, as similar levels between ISS and control group were observed.

Chemokines are cardinal in the pathogenesis of inflammation and the majority of chemokines are induced under pathological conditions by proinflammatory stimuli [29], that act firstly by increasing the level of mRNA. This increase, however, does not often reflect the circulating levels of protein, because some interferences can occur from transcriptional to translational processes. This phenomenon could explain our results on CXCL9 circulating concentrations.

Data regarding the levels of CXCL9 are limited and contradictory [20, 30, 31]. We observed an unexpected result, in fact the levels of CXCL9 were lower in pubertal subjects with respect to prepubertal of the same group, both in ISS and controls, suggesting a pubertal modulation. An age-related deregulation of immune system has been extensively reported in studies on aging, demonstrating a direct correlation between age and serum levels of CXCL10 and suggesting an enhanced Th1 immune response in aging. However, the correlation with age has not been shown for other chemokines [20, 32] and, although our evidence on pubertal modulation of CXCL9 is intriguing, no data are present in literature about this aspect. More studies are needed to elucidate or support this evidence.

In conclusion, this study describes two new genes (CXCL9 and FCGR1A/CD64) differently regulated in ISS and emphasizes an important involvement of IFN-γ, likely on regulation of cell cycle. In vitro studies are needed to validate this suggestion and understand how IFN-γ exerts its effects.

In our opinion, our data provide a new perspective for the investigation of ISS and associated implications.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer Science+Business Media, LLC 2012