Genetics of Waardenburg Syndrome in Africa: A Systematic Review

No.

Reviewer  Comments

Responses

Lines

Reviewer  1

1.

Why include participants mean age

Age of WS onset is key, since some associated clinical symptoms and/or features, such as hearing loss, can be age dependent. Age of affected individuals can also impact the rehabilitation prognosis of hearing loss.

2.

P.33 – 34, Why include reported BDP1 and MYO6 in the abstract.

Though the BDP1 and MYO6 reported may require further validations and additional report from unrelated  families/populations, to confirm a possible phenotypic expansion,  we think it is an important report that needs to be mentioned, to enhance targeted investigations of these genes in WS by other authors.

Lines 33-6

3.

P. 42 contradicts to P. 31

The statement has been revised accordingly:

This review showed that WS2 is the most common in Africa. Variants in PAX3 and SOX10 were the predominant genetic causes. The study emphasizes the need to further investigate in-depth clinical characterization, molecular landscape and the pathobiology of WS in Africa.

Lines

41-42

4.

It would be interesting to know how many PAX3variants have been described.

PAX3 variant  are listed in both Table 2 and Table S5.

PAX3 associated variants were frequently reported in 46.1% (12/26 families),

Manual curation of the reported variants revealed possible mutational hotspots at specific amino acid positions for some of the identified PAX3 and SOX10 variants in WS individuals. For example, at PAX3 (NM_181458.4):c.142G>T-p.(Gly48Cys) variant position, four previously reported pathogenic variants (p.Gly48fs, p.Gly48Ala, p.Gly48Arg, and p.Gly48Ser) that alter the same amino acid (a different nucleotide change in the codon), were identified (Table S5). In addition, PAX3 (NM_181458.4):c.808G>C-p.(Arg270Gly), known WS1 pathogenic variant, affects a residue at which other pathogenic substitutions have been descried in global cohorts, including p.(Arg270Cys), p.(Arg270His), p.(Arg270Pro), and p.(Arg270Ser). At the amino acid position 271, c.811C>T-p.(Arg271Cys) identified in a WS1 case corresponds to the same position as other known PAX3-p.(Arg271Gly) and p.(Arg271His) pathogenic variants which were previously described.

Lines

243-244

Lines 266-275

5.

P. 201: High proportion of unspecified cases, very interesting, but what does this mean? Is it reflect Poor description of the phenotype? These articles should be excluded from the analysis

Articles with poor case descriptions were already excluded from the final analysis  as stated in the methodology (figure 1), and discussed as follows:

The large fraction of participants lacking formal audiometry likely reflects systemic challenge, a severe shortage of audiologists, limited healthcare access, inadequate diagnostic equipment, and financial constraints across many African settings. [50].

Lines

338 -341

6.

Fig. 1 Tab 1 and 2 Fig. 1 should be removed to the appendices.

Figure 1, which is the PRISMA standard information retrieval chart is required in main body, per the journal editorial policy.

Table 1 and 2 also details key findings and parameters (clinical and genetic) which we believe is important for the readers  in this review. Therefore we will be grateful if the reviewer to tolerate their inclusions in the main text.

Line 158

Reviewer 2 Comments

1.

The wording "genetic profile" is vague and usually refers to somatic mutational signatures of cancers; "

Thank you for the suggestion.

We have amended our title as follows:

Genetics of Waardenburg Syndrome in Africa: A Systematic Review

We have also amended the text, where appropriate.

Line 2  

2.

Variant nomenclature should follow current HGVS guidelines: the RefSeq reference sequences (transcript for cDNA change and protein for aminoacid change) have not been stated anywhere in the text and in at least one instance (variant "c.587-?-?" in Table 2) the description is unintelligible.

Thanks for the suggestions. We have revised accordingly.  

For the specific variant the reviewer  has referred to, it was reported as a large deletion of exon 5 to exon 9, with no transcript information.

Table S5

and whole document

3.

These are variant (are they all SNV?) types. No "function" here.

Corrected, as follows:

Genes variants were missense (27/43), deletion (7/43), splicing (5/43), nonsense (2/43), indel (1/43), and duplication (1/43),

Lines  

37-38

4

I don't understand the logical connection behind this sentence.

We have simplified as follows:

There was no functional data to support the pathogenicity of putative causative variants.

Lines

39-40

5.

most common what?

Revised accordingly:

WS type 2 (WS2), with characteristically no dystopia canthorum, is the predominant subtype (36.9%; n = 31/84).

Line

30-31

6.

Are you sure that the prognosis for an individual patient can be reliably inferred from genotypes?

We agree, revised as follows:

Over 400 WS-associated variants have been reported [5], and multiple genetic subtypes have been defined, particularly within WS2 (WS2A, B, C, D and E) and WS4 (WS4A, B, and C). Nevertheless, the molecular underpinnings driving the variable clinical expressions and incomplete penetrance remain unclear.

Lines

60-64

7.

"most informative" = full clinical expression? "participant" = family member / individual per family?

There are only two families, from Ghana and Cameroon, with three and four members respectively. This makes a total of only six additional individuals. Perhaps it would have been better to include them separately in the cohort (otherwise intrafamilial phenotypic variability cannot be

The number families investigated has been added and highlighted in Table 1, showing participants family numbers.

Line 195

Table 1

8.

77 individuals in 78 families? One of the patients belongs to two families, then. Unless I am missing something. Please clarify.

 Families  numbers are now updated:

The 15 articles that met the study inclusion criteria described 84 WS cases from 57 families across ten (10) countries in Africa. Table 1 describes the WS participants reported in this article.

Lines

167-169

9

Singleton-WES or trio-WES? Agnostic or panel-based tertiary analysis?

Details provided as follows:

WES investigations typically involved sequencing at least two affected individuals per family, together with first-degree relatives, followed by Sanger sequencing to confirm candidate variants and perform segregation analysis.

Lines

245 - 247

10

What analytical method is used for TGS?

Is it perhaps targeted next-generation (multigene panel) (re)sequencing?

Sanger and TGS are reported as different methods both in the body of the text and in Figure 3. However, “targeted sequencing” does not unambiguously identify a method of analysis and could also refer to automatic bidirectional 'Sanger' SINGLE-GENE sequencing itself.

Details provided, as follows:

TGS approaches employed a 113-gene hearing loss panel sequenced on Illumina HiSeq 2500 platform.

Line 248

11

In Table 2, I count 34 cases that lack molecular characterization. 77- 34= 43 (not 44). Furthermore, one of the reported variants has been classified as benign (so 43- 1= 42 molecularly defined cases?).

Please reconcile discrepancies or clarify.

Thanks for you detailed revision. We are now corrected the discrepancies, as follows:

Reported variants classification established 44.2% (19/43) as pathogenic, 30.2% (13/43) likely pathogenic, 4.6% (2/43) benign, and 20.9% (9/43) variant of uncertain significance (Table 2).

Lines  281-283

12

Why 39? Where does this denominator come from?

The numbers have been fully revised.

13

What number? How many cases / families are solved / unsolved?

Perhaps I missed the part where you clearly stated this information.

The text has revised the text for  clarity and flow follows :

South Africa accounted for the largest proportion of WS cases in studies reviewed with 38.1% (32/84), followed by Tunisia with 26.2% (22/84) individuals (Figure 4). Putative Genetic etiology was reported in 71.4% (60/84) individual WS cases. Pathogenic variants were identified in four established WS genes, in a total of 26 families i.e. PAX3 (13 families), SOX10 (7 families), EDNRB (4 families), and EDN3 (1 family) were reported in Morocco, Tunisia, and South Africa respectively. Additionally, a candidate variant in PAX8 was described in one family from Ghana (Figure 4; Table 2). Thus, PAX3 associated variants were frequently reported in 46.1% (12/26 families), followed by SOX10 variants at 23.1% (6/26 families), and 15.3% (4/26) and 3.8% (1/26 families) in EDNRB, and EDN3 respectively. In addition, variants in two non-syndromic hearing loss (NSHL) genes (BDP1 and MYO6) were reported in two independent South African families with WS features, suggesting a possible phenotypic expansion.  A total of 24 cases remained unresolved:  including thirteen singleplex cases, two cases from one multiplex Ghanaian family, and nine cases from five multiplex South African families, even after WES investigations.

Lines

237-250

14

how many?

The detailed numbers are now  provided.

15

What does “neurogenesis disfunction” mean? The clinical signs themselves of W. syndrome are believed to be the result of a neural crest cell migration and specification defect.

This has been revised accordingly:

However, none of the neurological disorders common in patients with SOX10pathogenic variants were reported in individuals with WS4 we reviewed.

Lines  395-396

16

I do not understand what you mean by "distinct". It seems to me that the molecular epidemiology of W. syndrome that you have described in Africa is similar to what has already been observed in other areas of the world.

We have revised for clarity and flow as follows:

In addition, we found recurrent population/country specific variants e.g. in PAX3 (Table 2) suggesting possible founder effects that merit further investigations. If confirmed with larger sample sizes, such findings could enable implementation of an affordable targeted variant diagnosis strategy in clinical practice.

Lines

260-261

Lines

371- 374

17

I am sorry, but I am unsure about the logical link. It seems more like a mere conjecture.

This statement could only be made if a large proportion of 'bona fide' cases remained molecularly unsolved after a best-practice diagnostic pathway. However, earlier in the paper you demonstrated that this is not the case (i.e., incomplete clinical description, unavailability of molecular diagnosis, incomplete diagnostic pathway in the majority of case reports / series).

We have now removed the  statement.

Thank you.

No.

Reviewer 2 Comments

Responses

Line/Page

1.

broken sentence…

Corrected the sentence, now reads:

While there is a fair amount of literature on WS from other continents, a comprehensive study of WS in Africa has not yet been reported.”

Lines 73 – 74

Page 2

2.

Three out of the four pictures below apparently depict BLUE irides, not heterochromia iridum

Figure 2, sub-heading for panels c-f, and descriptions have been revised.

“Spectrum of eye pigmentation”. and “Panels (c), (d), and (f) show bilateral blue eyes (hypochromic blue). Panel (e) shows unilateral complete heterochromia iridum (brown right eye and blue left eye).”

Lines 202 – 212

Page 13

3.

urgen…..

Error corrected …urgent

Line 336

Page 21

4.

locus heterogeneity? (genetic h. is the overarching term for allelic h. & locus h.)

Corrected…now reads:

“Genetic characterization of WS-associated variants are reported in just a few countries (10/54). There was, as expected, a high allelic and locus heterogeneity for established WS genes”.

Lines 340 – 342

Page 21.

5.

I am not convinced of this "WS genes yet to be discovered" theory. The vast majority of "molecularly unsolved" cases is represented by purported WS2 diagnoses, where specificity of clinical features (prelingual SNHL + pigmentary anomalies, the latter term refers to a quite broad category) is by definition lower than in WS1/WS3/WS4.

That said, regardless of opinions, alternative explanations (alternative to the explanation of unknown causal loci) should be mentioned, e.g.:

- differential diagnosis (you just said that there are "substantial gaps in clinical documentation": maybe some of these poorly documented clinical hypotheses are wrong?): phenocopies, compound/blended phenotypes (dual/multiple diagnoses - mendelian or not),

- variants in known genes that go undetected using conventional first-tier tests like Sanger sequencing or capture/PCR-based short-read NGS (you discuss this later on, but it is not clear to me how these two points fit together).

We agree with the reservations of the reviewer and have discussed as follows:

However, the high proportion of molecularly unsolved cases, particularly among individuals labelled as WS2—where the defining clinical features (prelingual sensorineural hearing loss and pigmentary anomalies) are inherently less specific than those of WS1, WS3, or WS4—likely reflects a combination of diagnostic imprecision, as attested by substantial gaps in clinical documentation; phenotypic overlap with other hearing loss entities; blended or dual genetic diagnoses (Mendelian or otherwise); and pathogenic variation in known WS genes that remains undetected by conventional first-tier testing strategies such as Sanger sequencing or capture/PCR-based short-read NGS.

Lines 343 – 350

Page 21.

6.

This sentence could partly contradict your previous "high allelic heterogeneity" statement. In mathematical terms, allelic heterogeneity (AH) is the "maximum proportion of pathogenic variation within a gene attributable to a specific single allele". Although there is no standard threshold to define AH "high" or "low", maximum (100%) AH means that there is no recurrent variant (pathogenic alleles are all "private"). Founder variants, on the other hand, imply very low or no AH in that population.

The statement refers to specific recurrent variants identified in defined geographic settings, such as a PAX3 variant observed in multiple unrelated families in South Africa.

However, outside these localized contexts, high allelic heterogeneity remains the prevailing pattern in the other populations and settings reported in this study.

We have added the latest statement in the discussion.

 Lines 390-392

Page 22