Turner Syndrome Cytogenetics: A Complete Guide for Technologists

Published 2026-08-07

The 14-year-old who rewrote the rulebook

London, 1959. A 14-year-old girl is short for her age, has not started puberty, and has a webbed neck that her doctors have seen described in textbooks but never explained. Someone takes a bone marrow sample.

Charles E. Ford and his colleagues drop colchicine on the culture, freeze the cells mid-metaphase, and start counting. Three years earlier the field had finally settled on 46 as the human diploid number, so the counting part was new and the answer was supposed to be boring.

They count 45.

One X. No second sex chromosome anywhere. That single slide made Turner syndrome the first human sex chromosome abnormality ever identified, and it turned a vague clinical syndrome into something a technologist could see down a scope. Sixty-seven years later you are still counting those cells, just with better trypsin and a lot more rules.

Here is the part that still surprises people: 45,X is the only complete chromosomal monosomy compatible with postnatal human life. It shows up in roughly 1 in 2,500 to 3,000 live female births. It is also brutally lethal in utero, with an estimated 99% of 45,X conceptuses lost in the first trimester. The leading explanation for that surviving 1% is hidden somatic mosaicism: a normal cell line tucked away in some tissue that rescued the embryo. Keep that idea in your pocket, because it drives almost every counting rule later in this article.

Who described what, and when

The clinical picture arrived long before the chromosomes did.

The condition is still sometimes written as Ullrich-Turner syndrome, which is a fairer name than the one that stuck.

The Barr body detour

While the karyotype people were learning to count, the chromatin people were solving the same puzzle from the other end.

The buccal smear is obsolete as a diagnostic, but the tissue choice was ahead of its time. You will see buccal cells come back at the end of this article for exactly the reason Barr picked them: they are easy to get and they are not blood.

Why two patients with the same diagnosis look nothing alike

Turner syndrome has famously variable expressivity. Three things drive it:

1. The exact karyotype
2. How much mosaicism there is, and in which tissues
3. Whether the remaining X is structurally intact

That is not trivia. The indications written on the requisition decide how many cells you count and which probes you pull. So the clinical picture is a bench-workflow document, not just background reading.

The organ-by-organ summary

The SHOX story in one paragraph

SHOX (Short Stature Homeobox) sits in pseudoautosomal region 1 (PAR1) on distal Xp and Yp. PAR1 escapes X-inactivation, so a euploid person expresses two working copies. Lose the second sex chromosome, or delete distal Xp structurally, and one copy has to do the job of two. Chondrocyte proliferation suffers, and the whole skeletal spectrum above follows from that single dosage problem.

Analytical red flag: severe intellectual disability, autism spectrum traits, or profound developmental delay in a patient with a Turner-like phenotype points hard at a small ring X that lost its X-inactivation center. That case is in the structural section below, and it is the single highest-yield concept in this article.

Worth saying plainly: with educational and psychological support, the large majority of adults with Turner syndrome are fully employed and living independent lives. Elevated rates of anxiety, obsessive behaviors, social isolation, and low self-confidence are real and documented, and they are not the whole story.

Dosage compensation, or why losing one X matters at all

Here is the question that trips up most students: if a normal female silences one X anyway, why does having only one X cause anything?

First, the normal machinery.

X-inactivation (lyonization) was proposed by Mary Lyon in 1961. In any somatic cell with more than one X, all but one are epigenetically silenced and condensed into the Barr body. Which X gets silenced is normally random, cell by cell, early in embryogenesis, and the choice is then inherited clonally by every daughter cell.

The answer: inactivation is not absolute

Roughly 15% to 20% of X genes escape lyonization. They cluster at the distal ends of both arms, especially in PAR1 and PAR2, with others scattered along the chromosome. In a 46,XX female those genes are transcribed from both the active and the inactive X. A 46,XY male matches that dose using homologous copies on the Y.

So a normal human needs two functional copies of every escape gene. A 45,X individual has one. That is haploinsufficiency, and every feature above, from SHOX short stature to lymphatic and gonadal dysgenesis, is a downstream consequence of that deficit.

The karyotypes you will actually see

Classic monosomy X (45,X)

Non-disjunction is the failure of homologues to separate in meiosis I, or of sister chromatids in meiosis II. Either way you get a nullisomic gamete with no sex chromosome; fertilize it with a normal single-X gamete and you have a 45,X conceptus.

RFLP and microsatellite studies show the retained X is maternal in 60% to 75% of classic cases. Translation: the error usually happens in paternal spermatogenesis, whether by non-disjunction or anaphase lag with loss.

Two consequences that get tested:

• No advanced maternal age effect, unlike trisomy 21 or 18
• No established environmental risk factors

45,X/46,XX mosaicism

Mosaicism means two or more genetically distinct cell lines from one zygote. Start with a normal 46,XX fertilization, lose an X to anaphase lag during early cleavage, and a 45,X lineage grows up alongside the original cells.

The phenotype is generally milder: fewer left-sided heart defects, less metabolic syndrome and obesity, and a meaningfully higher chance of spontaneous menarche and natural pregnancy.

45,X/46,XY mosaicism, the one you go looking for

A 46,XY zygote loses the Y in a subset of cells. The patient presents as a phenotypic female with dysgenetic streak gonads, and there is hidden Y material in the mix.

This is why "rule out Y" is not an optional add-on for a Turner workup.

Structural X abnormalities and the failsafe

When the second X is structurally abnormal, biology stacks the deck: skewed, non-random X-inactivation preferentially silences the abnormal X and keeps the normal one active, minimizing imbalance.

Isochromosome Xq: 46,X,i(X)(q10)

The most common structural variant, 15% to 24% of all Turner karyotypes. An isochromosome is a mirror-image chromosome of two identical arms sharing a centromere, formed by transverse (not longitudinal) centromere misdivision or a U-type sister chromatid exchange in proximal Xp.

Ring X: 46,X,r(X) and the XIST question

Rings form when both telomeric ends break and the repair machinery fuses the sticky ends into a circle. Rings are mitotically unstable: sister chromatids tangle at anaphase, the ring breaks or is lost, and a secondary 45,X line often appears.

Everything about severity comes down to one thing: did the ring keep the Xic and a functional XIST?

That second row is not classic Turner syndrome, and mistaking one for the other misroutes the whole clinical evaluation.

Small supernumerary marker chromosomes (sSMC)

Sometimes G-banding shows 46 chromosomes: a normal X plus a tiny unidentifiable extra. Origin decides everything.

You cannot resolve this by banding. It takes FISH with centromeric probe cocktails or microarray to map the material.

On the bench: specimen to slide

G-banded metaphase karyotyping is still the gold standard here, because it shows numerical aneuploidy and macroscopic structure in one view. It only works if the pre-analytical side is clean.

The processing sequence:

Prenatal specimens are amniocytes or chorionic villus sampling. The pre-analytical trap there is maternal cell contamination (MCC), where maternal decidual cells get cultured alongside fetal cells. Suspect MCC when a 46,XX line appears in an otherwise 46,XY culture, or whenever you are assessing a possible 45,X/46,XX mosaic. Document it, and consider QF-PCR or another molecular assay to confirm fetal origin.

The counting rules, which are the whole ballgame

Statistical power comes from cell count. Since G-banding is labor intensive, ACMG publishes minimums by indication.

Routine constitutional analysis

Detection sensitivity by cell count, 95% confidence

Turner syndrome is defined by mosaicism risk, so the routine 20 is not enough. Trigger an extended mosaicism workup when a sex chromosome abnormality is clinically suspected, or when even one 45,X cell turns up during a routine 20-cell count. Plan on 30 to 50 metaphases minimum.

Clonality: real mosaic or slide-making artifact?

Slide making is violent. Membranes rupture early, chromosomes drift off, and you get artificial monosomies. That is pseudomosaicism, and the clonality rule exists to filter it out.

Turner syndrome is loss, so you need three 45,X cells before you report 45,X mosaicism.

In prenatal amniocyte cultures, which grow as discrete adherent colonies, add the multiple colony rule: abnormal cells must appear across independent vessels or colonies. Several 45,X cells all sitting in one colony in one dish is an in vitro event, not the fetal genome.

When the blood karyotype is normal but the patient is not

Peripheral lymphocytes come from mesoderm, and the hematopoietic compartment is under heavy selective pressure. A 45,X line can be depleted or wiped out of blood over time while thriving in other germ layer derivatives.

So a convincing phenotype (severe short stature, streak ovaries, high FSH) with a clean 50-cell blood karyotype does not close the case.

The payoff number: interphase FISH on uncultured buccal cells found that up to 29% of patients diagnosed with pure monosomy X by blood karyotype were actually tissue-specific mosaics with substantial 46,XX populations in buccal mucosa.

Molecular add-ons and what each one misses

G-banding sees down to roughly 5 to 10 Mb. Everything below that needs help.

FISH specifics

Workhorse probes: DXZ1 (X centromere alpha-satellite), DYZ3 (Y centromere), and SRY on distal Yp. Prefix ish for metaphase spreads, nuc ish for uncultured interphase nuclei.

Two high-yield uses:

• Metaphase FISH with a centromeric probe cocktail on an sSMC, to separate the gonadoblastoma risk from the intellectual disability risk
• Dual-color interphase X and Y FISH on buccal cells, scoring hundreds of ectodermal nuclei to rule out a hidden 46,XY line

CMA specifics

Pure 45,X reads as a continuous single-copy state across the whole X, with absence of heterozygosity (AOH) on SNP platforms. CMA is unbeatable for mapping exactly which genes an i(Xq) or r(X) gained or lost.

But know the limit cold: a 45,X/46,XX mosaic at 10% can be completely missed by microarray and caught easily by a technologist doing a 50-cell count or a 200-nucleus interphase FISH screen. That is your value on the bench, stated numerically.

Why NIPT false-positives cluster here

NIPT sequences cell-free fetal DNA from apoptotic placental trophoblast in maternal blood, maps millions of short reads, and flags read-depth deviations. For Turner it usually reports "high risk for X chromosome reduction." False positives are notably more common than for autosomal trisomies.

Any positive NIPT for Turner syndrome requires confirmation by amniocentesis or CVS, then karyotype, interphase FISH, or CMA.

ISCN 2020 quick reference

ISCN is the standardized language that gets your findings from the microscope to the clinician intact. Total count, comma, sex chromosome complement, then structural abnormalities. Mosaic clones are separated by a forward slash, listed in decreasing frequency, with cells analyzed in square brackets.

Karyotypes

FISH

Microarray

Prefix arr, genome build in brackets, chromosome and bands, nucleotide breakpoints, copy number state.

Five things to walk away with

Ford counted to 45 on one slide and opened a field. The tools got better, the resolution improved, and NGS reads pile up faster than anyone can review them. The technologist is still the link between raw genomic data and an actual patient, and that has not changed since 1959.

← Back to all CruxSci blog posts