a, At E14.5 no significant endogenous apoptosis (TUNEL-staining [green]) is observed in control forebrains (ctrl) whereas Lig4^Y288C embryos show increased apoptosis in the VZ/SVZ and IZ/CP with greater apoptosis being observed in the IZ/CP; representative sections of two Lig4^Y288C embryonic forebrains from different litters are shown (I, II). Tuj1 (red) and Tbr2 (red) antibodies are used to visualize the differentiated IZ or the intermediate precursors of the SVZ, respectively.
b, Costaining of the G2/M-marker pH3 (green, antibody from abcam) with Tuj1 or Tbr2 (red) in ctrl and Lig4^Y288C samples. Tuj1 labels the postmitotic IZ/CP compartment whereas the intermediate precursors of the SVZ are contained in the mitotically active compartment.
c, A dose-dependent increase in apoptosis is observed in the VZ/SVZ and IZ/CP of the forebrains of ctrl embryos 6h post IR at E14.5. Greater apoptosis is observed in the VZ/SVZ compared to the IZ/CP. Analysis following exposure to 3 Gy was from a single embryo. All other analysis, here and elsewhere, was from ≥ 2 embryos.
d, Quantification of apoptosis in the VZ/SVZ and IZ/CP of forebrains either untreated (− IR) or 6h after 0.1 Gy (+ IR) at E14.5. The % apoptosis represents TUNEL positive cells per 100 nuclei. Quantification was from 4-8 sections from 2 embryos representing 20,000-40,000 nuclei per condition. All quantification given was based on zones differentiated using pH3 staining. Results represent the average % apoptosis for the two embryos; error bars represent SD of the mean between embryos.
e, Lig4 ^Y288C embryonic forebrains are hypersensitive to 0.1 Gy (6h after IR at E14.5) in the VZ/SVZ and IZ/CP regions compared to the littermate ctrl forebrain (TUNEL-staining [green]). Additional concomitant Tuj1 (red) or Tbr2 (red) stainings are shown to distinguish the IZ from the SVZ regions. IR-induced apoptosis in Lig4^Y288C embryos is more than additive of the endogenous apoptosis plus the IR-induced apoptosis observed in ctrl embryos demonstrating hypersensitivity of both regions to apoptosis due to LigIV deficiency (see quantification in (d)).
f, Apoptosis (TUNEL-staining [green]) is highly elevated in the VZ/SVZ and IZ/CP of Lig4^Y288C embryonic forebrains 14h after 0.1 Gy IR at E14.5. Apoptosis in Lig4^Y288C IZ/CP is greater at 14h compared to 6h (additional costaining with Tuj1 (red)); apoptosis in ctrl embryos is less than at 6h. Thus, the relative difference is greater at 14 versus 6h. a,b,c,e,f: representative images of sagittal sections counterstained with DAPI. V = ventricle, scale bar 100μm.

a, Immunofluorescence (IF) staining of 53BP1 in E14.5 embryonic forebrains shows a gradient from low to high expression from the VZ at the ventricular surface towards the IZ/CP. The gradient was reproducibly more obvious in Lig4^Y288C mutant embryos (no IR) and after irradiation (0.1 Gy 1h, 0.5 Gy 1h) in ctrls compared to untreated ctrl (no IR) embryos. The stronger evidence of 53BP1 staining in the presence of DSBs is also observed in cultured cells and appears to represent foci formation rather than protein upregulation. The representative images are of sagittal sections: 53BP1 (red) and DAPI (blue) and overlay images are shown, scale bar 100 μm), areas of the IZ/CP and VZ/SVZ which are maginified in b) and c) are highlighted.
b, magnification of highlighted areas from panel a (I: IZ/CP area; II: VZ/SVZ area); overlay of 53BP1 staining (red) and DAPI (blue) demonstrates nuclear staining with 53BP1. The staining is pan nuclear but defined foci formation is evident after IR particularly in the IZ/CP. Whereas untreated ctrl forebrains show a low number of 53BP1 foci in the VZ/SVZ and IZ/CP, an elevated level of 53BP1 foci can be detected 1h after IR (0.1 and 0.5 Gy). 53BP1 foci numbers are greater in the IZ/CP region. The foci were difficult to count in the VZ/SVZ region and only the IZ/CP has been accurately quantified (see c). It is likely that in the VZ/SVZ the 53BP1 foci are underestimated particularly after low doses due to the low 53BP1 expression. Untreated Lig4^Y288C embryonic forebrains show elevated 53BP1 foci numbers in the VZ/SVZ and IZ/CP compared to ctrl embryos. The foci numbers in the IZ/CP are comparable to 0.1Gy 1h; the numbers in the VZ/SVZ appear similar to (although slightly less than) the numbers observed in ctrl embryos exposed to 0.1 Gy.
c, magnification of highlighted areas from panel a (I: IZ/CP area; II: VZ/SVZ area) 53BP1 staining (red) only; quantification of 53BP1 foci/cell in the IZ/CP is given below panel cI . Number of embryos for quantification were: untreated ctrl and 0.5Gy, 4; 0.1 Gy, 2; Lig4^Y288C, 3. Scale bar 20μm.

a, Analysis of mitotic cells in the irradiated embryonic forebrain. Unirradiated ctrl embryos (unt) show abundant pH3 stained cells (green) in the layer adjacent to the ventricular surface; irradiated embryos show diminished pH3 positive cells 1h after IR (0.1, 0.5 or 1 Gy). E14.5, representative snap-frozen sagittal sections, DAPI-counterstaining. Scale bar 100 μm.
b, Quantification of mitotic index (MI) 1h following exposure to IR doses shown. The MI is normalised to untreated ctrl embryos (unt). n: number of embryos analyzed per data point. * p < 0.001 for untreated ctrl vs. 0.5 Gy 1h.
c, Time course for release of G2/M checkpoint arrest of ctrl embryos after exposure to 0.5 Gy. E 14.5, Representative prefixed sagittal sections, pH3 (red), DAPI-counterstaining. V = ventricle, scale bar 100μm. This section used anti-pH3 antibody from Millipore in contrast to data in panel (a). This antibody detects G2 phase cells (weakly staining) as well as mitotic cells (brightly staining).
d, Quantification of MI following exposure to 0.5 Gy relative to untreated ctrl embryos (unt) (unt and 1 h time point as in (b)). 53BP1 foci/ cell in the IZ/CP at the given time points are also shown; n: number of embryos analyzed per time point. Anti-pH3 antibody from Cell Signaling.
e, γH2AX foci formation in VZ mitotic cells endogenously and 6h post 0.5 Gy. Cells were co-stained for γH2AX and pH3. Figure shows the percent mitotic cells at the ventricular surface with 0, 1 or ≥ 2 γH2AX foci for unirradiated (ctrl, unt) and 0.5 Gy irradiated ctrl embryos (ctrl, 0.5 Gy 6h) and unirradiated Lig4^Y288C embryos (Lig4^Y288C, unt). n: number of embryos analyzed. Ctrl and Lig4^Y288C samples were littermate pairs. The inset shows a representative mitotic cell 6 hours after 0.5 Gy harbouring 4 γH2AX foci (pH3 (red), γH2AX (green) separately and as merged image).

Features of the VZ/SVZ and IZ/CP impact upon their response to DNA damage (a) and can explain the different apoptotic sensitivities of the VZ/SVZ and IZ/CP in embryonic brains after IR and in NHEJ-deficiency, respectively (b): The VZ/SVZ is rapidly proliferating whereas cells in the IZ are predominantly in G0/G1 phase. The VZ/SVZ/IZ incurs high levels of endogenous DSBs which correlate with VZ/SVZ proliferation rate and declines from the IZ to the CP. Hyper-sensitivity of ctrl embryos to IR exposure is due to exquisitely sensitive activation of apoptosis in the VZ/SVZ by ATM-dependent and –independent processes. In contrast, microcephaly in LIG4 syndrome and XLF-deficient patients arises as a consequence of several features (i) high DSB induction in the VZ/SVZ/IZ cells due to the rapid proliferation rate in the VZ/SVZ (ii) the DSB repair defect that allows the endogenous DSBs to persist (iii) an inefficient G2/M checkpoint that allows cells with persistent DSBs to transit from the VZ/SVZ to the IZ (iv) high sensitivity of the IZ to persistent DSBs. The steady state level of unrepaired DSBs in LIG4 patients and Lig4^Y288C mice must lie below the level activating massive apoptosis or prolonged checkpoint arrest in order to generate a viable animal. These combined features explain the hypersensitivity of the embryonic brain to LigIV deficiency.

IR exposure causes directly induced DSBs (left panel), base damage and single strand (ss) breaks (SSBs). Replication past any of these lesions can cause replication fork stalling/collapse and ss regions of DNA (>25 nucleotides in length) (right panel). DSBs can also arise at collapsed forks. DSBs activate ataxia telangectasia mutated (ATM) signalling; ssDNA regions activate (ATM and Rad3 related (ATR) signalling following its recruitment to replication protein A (RPA) coated ssDNA regions of > 25 nucleotides. SSBs of a few nucleotides do not activate ATM or ATR. ATM and ATR signalling can activate checkpoint arrest and p53-dependent apoptosis. DSBs are predominantly repaired by DNA non-homologous end-joining (NHEJ), which uses DNA ligase IV as the ds DNA ligase (homologous recombination (HR) can also function in late S/G2). NHEJ involves binding of the Ku protein to ds DNA ends, recruitment of the catalytic subunit of the DNA-dependent protein kinase (DNA-PK), followed by end processing and recruitment of a ligation complex encompassing DNA ligase IV, XRCC4 and XLF. IR activates only ATM signalling in non-replicating cells (eg IZ) but ATM and ATR can be activated in replicating cells due to the generation of ss DNA regions at stalled replication forks. p53-binding protein 1 (53BP1) does not play a major role in ATR signalling but amplifies ATM signalling by aiding the tethering of ATM at DSBs. 53BP1 is also required for the repair of ~ 15% of DSBs (representing those that arise in heterochromatic DNA regions). γH2AX foci numbers closely respond to DSB formation in non-replicating cells, cultured cell models; 53BP1 is recruited to all DSBs and the numbers of 53BP1 and γH2AX foci usually closely correlate in non-replicating, cultured cells. The symbols are explained in the figure. Light circles on the DNA represent histones and darker circles represent RPA bound to ss regions of DNA.

a, No significant apoptosis was observed in unirradiated forebrains of ATM^+/− embryos.
b, Marked apoptosis is evident 6h after 0.5 Gy in ATM^+/− forebrains in the VZ/SVZ and IZ/CP and is markedly reduced in the VZ/SVZ of ATM^−/− embryos. No detectable increased apoptosis is observed in the ATM^−/− IZ/CP region. a,b: TUNEL [green], counterstain with DAPI, sagittal sections, E14.5, V = ventricle, scale bar 100μm.
c, Quantification of apoptosis in the VZ/SVZ and IZ/CP of forebrains from ATM^+/− or ATM^−/− embryos 6 hrs after 0.5 Gy. Apoptosis (%) represents the TUNEL positive cells per 100 DAPI nuclei; E14.5, n = 3 embryos per condition. One representative section per condition was quantified. In total ~7500 and 9500 ATM^+/− and ATM^−/− nuclei were counted, respectively, * p < 0.005 for VZ/SVZ and IZ/CP.

a, 53BP1 foci were enumerated in the IZ/CP and developing cortex of ctrl and Lig4^Y288C animals at E14.5, E17.5 and P5. Number of embryos (n) analyzed is given below the respective histogram column. * p < 0.001 for Lig4^Y288C P5 compared to E14.5.
b, Apoptotis (%) (TUNEL positive cells/100 nuclei) for Lig4^Y288C at E14.5 (as in Fig. 2d) and E17.5 is given. 2 embryos were analyzed per time point.
c, Images of 53BP1 (red) and TUNEL (green) staining are shown for the IZ/CP region of ctrl and Lig4^Y288C sections at E14.5 and for the Lig4^Y288C developing cortex at E17.5 and P5 (sagittal sections, 40 x magnification, insert 100 x magnification).

Pregnant mice were treated with a standard dose of BrdU (i.p.) at E14.5 and mock irradiated or irradiated with 0.5 Gy 30 min later. The embryos were collected at 1h and 14.5h post IR. 14.5h post IR was chosen as the time for analysis based on an estimated cell cycle time of 15 h in the VZ/SVZ at E14.5 (Mitsuhashi and Takahashi, 2009). 0.5 Gy was chosen as a dose that induced apoptosis but not prolonged checkpoint arrest. Exposure of embryos to BrdU for 30 min at E14.5 and harvesting 1h later revealed a single broad band of BrdU staining predominantly within the replicating VZ/SVZ region as assessed by Tbr2 staining (panel a). Examination at 15h post BrdU addition showed that BrdU staining was now also observed in the differentiated IZ compartment as assessed by Tbr2 and Tuj1 staining (panels b and c). Following exposure to 0.5 Gy IR at 30 min post BrdU addition and harvesting 14.5h later, 10-20 % of such cells were TUNEL positive.
a, Analysis of embryos exposed to BrdU for 30 min prior to 0.5 Gy IR and harvest after 1h. Section subjected to BrdU and Tbr2 double-staining. BrdU⁺ cells are predominantly located within the VZ/SVZ layer as defined by Tbr2 staining;
b, Analysis of unirradiated embryos exposed to BrdU for 15 h prior to analysis. Sections examined as in panel a. BrdU⁺ cells can now be observed in the IZ layer located beyond the Tbr2⁺ layer.
c, Analysis of embryos treated as in panel b) and sections subjected to triple Tuj1/BrdU/TUNEL staining. There is no significant level of apoptotic cells demonstrating that BrdU exposure does not activate apoptosis.
d, Analysis of embryos exposed to BrdU for 30 min, then treated with 0.5 Gy IR and harvested 14.5h later. Sections were subjected to triple Tuj1/BrdU/TUNEL staining.
e, magnification of the insert from the IZ zone in panel d) demonstrating the stainings separately (BrdU (red), TUNEL (green), DAPI (blue) and Tuj1 (grey)) and overlaid; two representative BrdU/TUNEL double⁺ cells are marked by arrows.
a-d, Representative areas of sagittal snap-frozen forebrain sections, a-b, single channel images in grey scale (DAPI, Tbr2, BrDU), overlay image of Tbr2 (green) and BrdU (magenta). c-d, single channel images in grey scale (DAPI, Tuj1, BrDU, TUNEL), overlay image BrdU (red)/ TUNEL (green)/Tuj1 (blue), a/b/e) deconvolved images, c/d) non-deconvolved images.