General comments
We appreciate this document as a successful achievement of a new attempt of ICRP to enhance robustness of the radiation risk assessment. The authors provide the hypothetical framework to explain the risk models used in the radiological protection (RP) system and epidemiological data as the basis of risk assessment to take the recent knowledge of stem cell biology. Especially, the concept of tissue-level quality control by stem cell competition would be very important to improve the RP system with taking into account of dose-rate effect and age dependency, which were not sufficiently considered in the current system. We hope that the viewpoint shown in this document would be reflected to the on-going discussion on DDREF.
In addition, the section 3.9. "recommendations for future research" is consistent with the initiative of "Recommending research needed to strengthen the System of Radiological Protection" shown in ICRP Strategic Plan 2011-2017.
Specific comments:
- Introduction
The significance of this document would become easier to be understood when it is explicitly stated that the "Bergonie-Tribondeau's law" alone cannot explain tissue radiosensitivities. For example, before the 4th sentence "Stem cell radiosensitivity varies both within and among tissues." in (6), insert following:
Classically, cellular radiosensitivity has been simply considered to be associated with the reproductive activity, according to law of Bergonié and Tribondeau,
- 2.3.5. (43) on Cairns' hypothesis
Cairn’s hypothesis (‘immortal strand hypothesis’), as well as stem cell competition, has an important role in this report to speculate the mechanism of the radioresistance of quiescent premature stem cells. However, we are afraid that it may be hard to comprehend the immortal strand hypothesis by quite a few readers. We suggest to add a figure, e.g. Figure 1 in Rando et al. (Cell 2007 (129) p1240).
- 2.5.5 (59)
In this section, a paper written by Bondar and Medozhitov was cited (Bondar and Medozhitov, Cell Stem Cell. 2010). They found that radiation-induced competition depends on the p53 level in the competing cells because HSCs from the untreated donor cannot be replaced a significant number of HSCs from irradiated p53+/- donor mice. Their evidences are not compatible with the previous sections of this document that damaged/mutated stem cells may be excluded from the tissue. It should be emphasized that p53-depletion is the critical malignant factor in multi-stage carcinogenesis. Furthermore, they transplanted p53-mutated HSCs into recipients 9-times higher than wild-type HSCs. Considering the outcompeting potential, it is easy to consider that mutated/damaged HSCs would be the winner of the stem cell competition. Generally, this situation does not apply to low-dose-rate exposure because the induction of p53 mutation on stem cells may be low and p53-mutated stem cells are very unlikely to outcompete undamaged stem cells under this condition.
- 3.1.3. (73)
An act of non-targeted effect, which includes bystander effects and genomic instability, in the low dose range was described. Recently we showed that low dose of soft X-rays could not induce a bystander response in normal human fibroblasts at doses < 100 mGy using X-ray microbeams (Tomita et al. J Radiat Res. 2012; 53(3):482-8, and Tomita et al. Radiat Res. 2010 Mar;173(3):380-5).
- 3.1.4 (81) (82)
There are no corresponding sections to the reference of "see section 3.6.2". Authors should state here or in section 3.6 about the discussion summarized in Executive summary H.
- 3.3.1 (91)
Need reference for "At the cellular level, there is ample evidence that the linear term of the mutation induction rate by radiation is independent of dose rate." Maybe the same references as the next comment.
- 3.3.2 (94)
The dose-rates studied in Russell and Kelly, 1982, which showed that the mutation induction rates was the same between 0.8 R/min and 0.001 R/min, were much higher than the dose-rate discussed here. Tanaka et al 2013, the dose-rate effect of dicentric chromosome aberrations in mouse lymphocytes less than 20 mGy/day, and Tucker et al 1998, the dose-rate effect of mutation frequencies in mouse intestinal cells among 55, 18.5, and 6.4 mGy/day, were more appropriate.
- 2.3.4 (41), 3.2.1 (86): on the "error-prone NHEJ repair"
In this report, a role of non-homologous end-joining (NHEJ) in a process of DNA double-strand break (DSB) repair in the non-cycling quiescent tissue stem cells was well described and discussed. For example, in section 2.3.4. DNA repair in stem cells, it was described that “Firstly, quiescent stem cells rely on NHEJ, but this repair pathway is considered to be error-prone.” (p38 (41) line 1509-1510). In addition, it was also described that “In addition, quiescent cells have to be dependent on the error-prone NHEJ repair for coping with DNA damage. Thus, the benefit of quiescence us dependent on the trade-off if avoiding replication-mediated mutation versus taking a chance of damage accumulation and resulting mutations.” in section 3.2.1 (p54 (86) line2176-2179). Many documents including this report have been suggested that HR is an error-free process, while NHEJ is an error-prone DSB repair pathway. However, it seems that these arguments should be revised from the current accumulating knowledge. Recent reports (reviewed in Bétermier et al. PLoS Genet. 2014 Jan;10(1):e1004086 and so on) showed an intrinsic precision of canonical C-NHEJ (C-NHEJ is also referred to as core-NHEJ or classical-NHEJ), which is Ku heterodimer, DNA-PKcs, XRCC4 and ligase IV dependent). On the other hand, alternative-NHEJ (A-EJ or alt-NHEJ, which is also referred to as backup-NHEJ) is suggested as a highly error-prone NHEJ. The Task Group should carefully discuss the fidelity of NHEJ taking the difference between C-NHEJ and alt-NHEJ into consideration.
-Annex A
In paragraph (77) (p50), it is described that “In contrast, leukaemia and childhood cancer require much fewer mutations such as two, which can be supplied by radiation alone.”. This sentence is very important for considering the mechanism of radiation leukemogenesis. However, there is no related suggestion and/or discussion in Annex A.
The Task Group should indicate the more grounds of this foundation by quoting some papers in Annex A.
- A.3.2. (A31)
@In section A.3.2. ((A31) p84, line3385-3388), it was described that “Further, in a report of cancer risks (solid cancers and leukaemias) within 8,000 children of female Mayak nuclear workers who were exposed in utero to low doses (mean doses of ~55 mGy) of ionising radiation, any significant elevated cancer risks were not shown (Schonfeld et al., 2012).”. This result is very important for considering the risk from fetal-stage exposures. This cohort study (Schonfeld et al., 2012) should be quoted and discussed in section 3.7.2. (A117) (p64 line2622-p65 line 2649).
- A.7. (A87)
It is inappropriate to show a specific number of "a factor of 10" as a result of comparison between mutation induction rate per unit dose and spontaneous rate.
- C.1. (C2)
New information after Richardson 2009 should be reflected for "little or no effect from adult exposures". Mabuchi et al 2013 reported that 3 among 9 studies of adult exposures showed the increase of risks.
- F.7.
This section should be deleted, because the contents of this section are not appropriate as "Mutagenesis".
Editorial comments:
- A.1.1. (A7): "Jablon and Kato, 1970" and "Court Brawn et al 1960" quoted here are not in the Reference.
- Table 2.1.: Species should be clarified.
- Fig A.7.: The positions of indicators a-d should be adjusted.
- A.4. (A48) line 3593, no ")" after "conditional medium"
- A.6. (A67) line 3857: Kit→c-Kit, line 3870: ptotein→protein
- A.7. (A74): typo "a.y"
- F.2.: The section numbers are irregular (two F.2.1.)