*3.2. Infections and Autoimmunity*

Recurrent upper respiratory infections were detected in the majority of the cohort, including otitis and sinusitis (75%); lower respiratory tract infections occurred in two patients (25%), requiring hospitalization. No patient developed bronchiectasis. A total of 37.5% (three out of eight subjects) were affected by recurrent periodic fevers, without genetic features of monogenic autoinflammatory syndromes, and subject P2 was affected by recurrent urinary infections. Subjects P2 and P3 presented with recurrent herpes simplex infections and aphthous stomatitis. No fungal or opportunistic infections were identified during the follow-up, except for one episode of oral candidiasis in patient P2, requiring fluconazole prophylaxis. Regarding severe infections, an episode of mediastinitis in the context of the sternal surgical wound was described in child P8, with a good outcome.

Concerning autoimmune manifestations, juvenile idiopathic arthritis with oligoarticular phenotype was diagnosed in patient P3. Joint involvement (the right knee, right temporomandibular, and proximal interphalangeal of the fifth finger of the right hand) required a step-up therapy with frequent evacuative and infiltrative (triamcinolone acetonamide) arthrocentesis, use of modifiers of the biologic response (methotrexate), and biologic drugs (anti-tumor necrosis factors). Moreover, the child was unresponsive and developed a chronic inflammatory process that resulted in deformity of the hand joint and residual synovial inflammation of the right temporomandibular joint, as detected using nuclear magnetic resonance, with functional limitation.


**Table 1.** Clinical features of the patients.

*J. Clin. Med.* **2022**, *11*, 2025

right-sided aortic arch.

#### *3.3. Auxological and Endocrine Features*

Weight and head circumference were within normal percentiles for the overall cohort, while a delay in height growth was remarked. No measurements were reported for subject P8 and, consequently, he was excluded. A reduction in the height (between −1 and −2 SD) was reported in almost all of the cohort, prevalently in childhood. No case of short stature (height < −2 SD) was registered. Of note, only the two sisters P4 and P5 presented within a height normal distribution (Supplementary Figure S1).

We focused on the endocrine features of the cohort by investigating the 25OHD level and its association with immunological parameters, as we previously published for a cohort of DGS patients [23]. Therefore, we identified three subgroups according to 25OHD levels: group A (normal values), group B (25OHD deficiency), and group C (patients with hypoparathyroidism). Toddler P8 was excluded due to the absence of 25OHD determination. Group B was composed of patients P2, P6, and P7. The vitamin D status was performed at diagnosis and showed the following 25OHD mean values: 30.23 ng/mL, 17.43 ng/mL, and 14.0 ng/mL in the A, B, and C groups, respectively (normal range < 20 ng/mL). Both the A and B groups presented with parameters of phospho-calcium metabolism (parathyroid hormone, PTH, calcium, phosphate, alkaline phosphatase) in the normal range for the patients' ages. Patients belonging to group B received oral supplementation with cholecalciferol at the dosage of 25,000 IU/month. An increase in 25OHD values was observed in P2 (46.9 ng/mL vs. basal value 16.8 ng/mL) and P6 (35.5 ng/mL vs. basal value 17.63 ng/mL).

Patients P4 and P5 (25%) received a diagnosis of congenital hypoparathyroidism. Both presented with neonatal hypocalcemia, elevated serum phosphorus concentration, PTH values constantly suppressed or undetectable (<3 ng/mL), and low vitamin D values (13 ng/mL and 15 ng/mL in P4 and P5 patients, respectively). They received treatment with calcium carbonate (50–100 mg/kg/day) and calcitriol (20–40 ng/kg/day) for at least 6 years, with good outcomes. Moreover, the vitamin D values returned to within the normal range only for a short time and subject P4 suffered from two hypercalcemia episodes during the follow-up.

However, we did not observe any difference between vitamin D levels and the immunological parameters examined in the various groups.

#### *3.4. Immunological Features*

The lymphocyte and lymphocyte subpopulation values of patients, compared with age-matched normal controls, are shown in Tables 2 and 3. Lymphocytopenia was reported in 62.5% of the cohort (five out of eight patients). A low absolute count of CD3+ lymphocytes and CD4<sup>+</sup> and CD8<sup>+</sup> cells were observed in 75% (*n* = 6), 87.5% (*n* = 7), and 75% (*n* = 6) of subjects, respectively. The B cell absolute number was low in two lymphopenic patients, with them being normal if expressed as a percentage. Similarly, the NK cell count exhibited the same behavior in patient P2; conversely, high relative numbers of NK cells were observed in P3, P6, P7, and P8. Furthermore, extensive phenotyping was performed due to the suspicion of a putative thymic a/hypoplasia. RTEs were lower than normal agematched values in six out of eight patients, in particular, P1, P2, and P3 < 15%. Naïve helper T cells were reduced in P2 and severely low in both patients P1 and P3 (8.8% and 12.7%, respectively). Naïve cytotoxic T cells were reduced in two subjects (P3, P6). Central memory cytotoxic T cells were low only in P8, whereas they were higher than the normal range in P1, P6, and P7. Both effector memory CD8<sup>+</sup> T cells (EM and TEMRA) were reduced in patients P5 and P7, while EM was only low in P4 and TEMRA only in P1. The overall cohort showed normal values of follicular T cells. Regulatory T cells were increased in patient P2 and reduced in patient P6. The analysis of the B cell compartment (naïve and switched memory cells) revealed mostly normal values, with the exception of two patients who showed a reduction of naïve B cells (P6 and P7) and an increase of switched memory cells (P6).

To explore the interface between innate and adaptative immunity, we investigated the dendritic cells, as shown in Table 4. A total of 50% of patients (P2, P3, P5, P7) exhibited a reduction in total DCs in comparison to healthy controls, whereas an increase was observed in patient P8. The observed reduction affected mainly the myeloid subset (37.5%), except for patient P2, who showed a reduction in both subsets.

Serum immunoglobulin levels are shown in Table 5. Two children (P2 and P6) had partial selective IgM deficiency (between −1 and −2 SD). Patients P3 and P5 (25% of the cohort) had a reduction in both IgG and IgA, with the IgA values being severely low (>−2 SD) in P5, whereas in P8, a reduction in both IgG and IgM (between −1 and −2 SD) was observed. P4 showed hypogammaglobulinemia (both IgG and IgA, between −2 and −3 SD), along with a compensatory increase (>+3 SD) in IgM levels, requiring immunoglobulin replacement therapy. Conversely, P1 had an increase in IgM (>+2 SD). IgE levels were normal in the overall cohort. The immunoglobulin subclass levels were in the normal range in the overall population, except for the IgG3 subtype, which was reduced in patient P8.

None of the patients had neutropenia. Mild eosinophilia (760/μL) was observed in subject P3.


**Table 2.** Lymphocyte subsets of the patients.

The absolute numbers of cell subsets are indicated for each patient (upper line). Lower lines indicate normal values for age (median (10–90th percentile)). \* Age at immunological evaluation; NK: natural killer.

cells (%) <sup>e</sup>


**Table 3.** Advanced phenotypes of the patients.

The frequency of cell subsets is indicated for each patient (upper line). Lower lines indicate normal values for age (median (10–90th percentile)). \* Age at immunological evaluation. <sup>a</sup> % of total peripheral lymphocyte population; <sup>b</sup> % of helper T lymphocyte population; <sup>c</sup> % of cytotoxic T lymphocyte population; <sup>d</sup> % of CD4+CD45RO+ T lymphocytes; <sup>e</sup> % of B lymphocyte population; NK: natural killer; TEMRA: effector memory T cells re-expressing CD45RA; CM: central memory; EM effector memory; RTE: recent thymic emigrants; Treg: regulatory T cells.


**Table 4.** Absolute and relative numbers of the dendritic cells in the cohort.

The frequency and absolute numbers of cell subsets are indicated for each patient (upper line). Lower lines indicate normal values for age (mean (10–90th percentile)). \* Age at immunological evaluation; <sup>f</sup> % of WBC; WBC: white blood cells; DC: dendritic cells; mDC: myeloid dendritic cells; pDC: plasmacytoid dendritic cells.

**Table 5.** Immunoglobulins and their subclasses in the cohort.


The frequency of immunoglobulins and IgG subclasses is indicated for each patient (upper line). Lower lines indicate normal values for age expressed as mean ± SD or 10–90th percentile; NA, not available. \* Age at immunological evaluation.

#### *3.5. Genomic Features*

The genetic assessment of all patients did not show any deletion in the 22q11.2 and 10p13-14 regions. Among the eight patients tested using array-CGH, two subjects did not show any pathological CNVs, according to the DGV database (http://dgv.tcag.ca/variation (accessed on 25 February 2022)), which collects variations reported in normal subjects. Table 6 shows the CNVs detected in the remaining six patients, along with the positions of the first and last abnormal probes and the extent and gene content of each CNV. The pattern of inheritance was assessed in four patients in which CNVs were inherited from the mother.


**Table 6.** Genetic abnormalities on array-CGH of the cohort.

OMIM genes are in bold; NA: not available; \* negative array-CGH.

Patient P1 showed a duplication of 314 kb in 2q24.1, inherited from her mother, that harbored the long intergenic non-protein coding RNA 1876. No overlapping duplications were reported in the Decipher database (https://decipher.sanger.ac.uk/ (accessed on 25 February 2022)), with this CNV classified as a variant of unknown significance (VOUS).

P4 and P5 were two sisters who showed a duplication in 20p11.22 inherited from their mother. This CNV has never been reported in individuals with a pathological phenotype (Decipher). This region harbors two non-coding protein LINC01727, LINC01726 and two coding protein PAX1 and NKX2 genes. PAX1 (\*167411) encodes a transcription factor implicated in embryogenesis in vertebrates and plays an important role in segmental spine formation and thymus organogenesis. In humans, the phenotypic effects of the PAX1 duplication have not been yet described, whereas homozygous mutations of this gene have been associated with otofaciocervical syndrome 2 with T-cell deficiency (OTFCS2) (#615560) [24–26]. NKX2 contains a homeobox domain. It is highly expressed in the central nervous system and encodes for a protein that is likely a nuclear transcription factor involved in the morphogenesis of this system. According to these data, this CNV is likely pathogenetic.

In patient P6, two duplicated regions were detected in 11p15.5 and the pseudoautosomal region Xp22.33/Yp11.32. The duplication in 11p15.5 encompassed a total of 15 genes (Supplementary Table S1), among which, EPS8L2, TALD01, SLC25A22, PNPLA2, and CD151 are reported in the OMIM database (https://www.omim.org (accessed on 25 February 2022)). Apparently, none of these five genes appeared causative of the clinical features of this patient (Supplementary Table S1). Moreover, we highlight that the pathological phenotype described in OMIM is related to homozygous loss-of-function mutations. No overlapping duplications are reported neither in the literature nor in Decipher; thus, according to the few data available, this variant could be classified as VOUS. The duplication in Xp22.33/Yp11.32 harbors one non-coding gene (LINC00685) and four coding genes, including SHOX (\*312865/\*400020) (Supplementary Table S1). SHOX/SHOX enhancer deletions cause short stature and skeletal abnormalities (#249700, #127300, #300582); microduplications in the pseudoautosomal region including SHOX appear to be rare and have been related to autism spectrum disorders and neurodevelopmental pathologic conditions [27].

Patient P7 showed a deletion in 17q21.31 of about 493 kb that was causative of Koolen– De Vries syndrome (KDVS) (#610443). Moderate-to-severe intellectual disability, hypotonia, and a characteristic face represent the core phenotype of this syndrome. More variable features include cardiac, genitourinary anomalies, seizures, nasal speech, and a friendly

demeanor [28]. Haploinsufficiency of KANSL1 (\*612452) appears causative of the syndrome, as the clinical phenotype does not substantially differ between patients with 17q21.21 microdeletion encompassing KANSL1 and patients with a de novo heterozygous mutation in this gene. KANSL1 encodes a nuclear protein that plays a role in chromatin modification. It is a member of a histone acetyltransferase (HAT) complex [29].

Patient P8 showed two deleted regions in 17q13.2 and Xq24, with sizes of 258 kb and 68 kb, respectively, both inherited from his mother. The deletion in 17p13.2 harbors WSCD1, which encodes for a membrane protein with a sulfotransferase activity, whose role is not yet known. This CNV can be classified as a VOUS. The deletion in Xq24 encompasses two genes, namely, CXorf56 and UBE2A, and the phenotypic effects of their loss-of-function in males are reported in OMIM. CXorf56 is related to "Intellectual development disorder, X-linked 107" (#301013). UBE2A encodes a member of the E2 ubiquitin-conjugating enzyme family that is required for post-replicative DNA damage repair and may play a role in transcriptional regulation; its loss-of-function is causative of "UBE2A deficiency syndrome" or "X-linked Nascimento-type intellectual disability syndrome" (#300860). According to these data, this CNV is pathogenetic.
