- Open Access
The effect of radio-adaptive doses on HT29 and GM637 cells
Radiation Oncologyvolume 3, Article number: 12 (2008)
The shape of the dose-response curve at low doses differs from the linear quadratic model. The effect of a radio-adaptive response is the centre of many studies and well known inspite that the clinical applications are still rarely considered.
We studied the effect of a low-dose pre-irradiation (0.03 Gy – 0.1 Gy) alone or followed by a 2.0 Gy challenging dose 4 h later on the survival of the HT29 cell line (human colorectal cancer cells) and on the GM637 cell line (human fibroblasts).
0.03 Gy given alone did not have a significant effect on both cell lines, the other low doses alone significantly reduced the cell survival. Applied 4 h before the 2.0 Gy fraction, 0.03 Gy led to a significant induced radioresistance in GM637 cells, but not in HT29 cells, and 0.05 Gy led to a significant hyperradiosensitivity in HT29 cells, but not in GM637 cells.
A pre-irradiation with 0.03 Gy can protect normal fibroblasts, but not colorectal cancer cells, from damage induced by an irradiation of 2.0 Gy and the application of 0.05 Gy prior to the 2.0 Gy fraction can enhance the cell killing of colorectal cancer cells while not additionally damaging normal fibroblasts. If these findings prove to be true in vivo as well this may optimize the balance between local tumour control and injury to normal tissue in modern radiotherapy.
It is widely accepted that the shape of the dose-response curve at low doses differs from the linear quadratic model . Induced radioresistance, hyperradiosensitivity or adaptive responses (i.e. a biopositive effect induced by a low priming dose and identified after application of a higher challenging dose) may occur at low doses of irradiation. The radio-adaptive response was first recognized 1984, when Olivieri et al. demonstrated that human lymphocytes exposed to low concentrations of radioactive thymidine show fewer chromatid aberrations caused by a 1.5 Gy challenging dose than those not pre-exposed to irradiation . Several publications have studied the effect with different cell lines, different pre-irradiation doses, and variable challenging doses [3–10]. However, the exact mechanism of the effect is yet unknown, thus precluding predictions whether a cell line will show an adaptive response or not. An altered gene expression caused by low-dose ionizing radiation has been identified. A radio-adaptive response seems to be associated with an up-regulation of DNA repair and stress response genes and a down-regulation of cell cycle control and apoptosis genes. TP53 (Tumour Protein p53) is supposed to play an important role in this mechanism . Protein synthesis, metabolism and signal transduction appear to be involved in the adaptive response as well . However, controversy remains regarding the mechanism and role of the adaptive response . This is probably due to cell-type and tissue-specific variations and different experimental conditions [12, 13].
Most radio-adaptive response experiments focussed on basic research of this phenomenon, with only few studies concentrating on its clinical applications, e.g. in radiotherapy .
We had previously described a difference between the reaction of normal bladder cells (HCV29) and that of bladder cancer cells (RT4) to different adaptive doses of irradiation. HCV29 cells showed an induced radioresistance after pre-irradiation doses of 0.05 Gy or 0.1 Gy, whereas RT4 cells displayed a hyperradiosensitivity after pre-irradiation with 0.05 Gy, 0.1 Gy, 0.2 Gy or 0.5 Gy .
While bladder cancer is only infrequently treated by radiotherapy, pre- or postoperative irradiation of stage II or III colorectal cancer is very common. These tumours are responsible for 655.000 deaths/year worldwide  with an incidence of 88.3/100.000 men and of 84.9/100.000 women in Germany in 2002 . Colorectal cancer is thus one of the most common cancers after prostate cancer for men and breast cancer for women. It is therefore of utmost importance to optimize the treatment for colorectal cancer in order to attain a high cure rate and minimized side effects. Radio-adaptive doses applied may probably aid to achieve this end as an adjunct to standard chemo-radiotherapy.
It was therefore our aim to evaluate the effect of different pre-irradiation doses followed by a 2.0 Gy fraction on a colorectal cancer cell line (HT29) and on normal fibroblasts (GM637).
Both cell lines were routinely grown in 80 ml flasks (NUNC, Wiesbaden, Germany). For HT29 cells the medium consisted of 83% McCoy's 5A medium supplemented with 16% fetal calf serum and 1% of a mixture of antibiotics (104 IU penicilline/ml and 104 μg streptomycin/ml). The medium for GM637 cells was a mixture of 82% minimum essential medium MEM (Eagle) with Earle's salts, 25 mM HEPES and without L-glutamine, of 16% fetal calf serum, of 1% sodium pyruvate 100 mM and of 1% of the antibiotic mixture (104 IU penicilline/ml and 104 μg streptomycin/ml). The cell lines were incubated at 37°C with 5% CO2, 95% humidity and a pH of 7.4. Cells were passaged in the exponential growing phase once a week, using 0.05% trypsin plus 0.02% EDTA in PBS at 37°C.
96-well culture plates were used for all experiments. Cells were seeded at a density of 250 cells per well (250 cells in 200 μl medium). Each plate contained wells with HT29 and wells with GM637, so that both cell lines were treated in the exact same way. Additionally, another plate was seeded with an increasing cell number per row (62.5-125-250-500-750 cells per well) for cell growth monitoring and survival reference.
After an incubation period of 24 h the plates were irradiated with 0 Gy, 0.03 Gy, 0.05 Gy or 0.1 Gy at a dose rate of 0.03 Gy/min (225 kV, 5 mA, 0.35 mm Cu). 4 h after pre-irradiation cells were further irradiated with 0 Gy or 2.0 Gy at a dose rate of 1.0 Gy/min (225 kV, 15 mA, 0.35 mm Cu). As a result, eight different irradiation groups were evaluated: 0 Gy (control), 0.03 Gy alone, 0.05 Gy alone, 0.1 Gy alone, 2.0 Gy alone, 0.03 Gy plus 2.0 Gy, 0.05 Gy plus 2.0 Gy and 0.1 Gy plus 2.0 Gy.
We chose the pre-irradiation doses to be 0.03 Gy, 0.05 Gy and 0.1 Gy respectively following an earlier study  that demonstrated pre-irradiation doses of 0.05 Gy and 0.1 Gy, but not of 0.5 Gy to be effective.
Cell viability test
The plates were incubated for an additional 7 days. The medium was subsequently removed from all wells. Cells were washed with PBS and 100 μl medium with 10% WST-1 (tetrazolium salt 4- [3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) were added to all wells. WST-1 is cleaved to a water-soluble formazan dye whose amount directly correlates to the number of metabolically active cells and is quantified spectrophotometrically by an ELISA reader at a wavelength of 450 nm (reference wavelength: 690 nm). The optical density was measured immediately (background measurement) and after 3 h.
All experiments were repeated three times resulting in at least 55 single data sets per irradiation group and cell line. The standard curve for control cells was checked to be certain that the cells of the experimental plates are in the exponential phase of the survival curve and not in the plateau phase. After subtracting the background the relative cell survival of all wells was calculated. Using the Student's t-test the statistical significance of the results (p ≤ 0.05) was evaluated.
HT29 cell studies
An irradiation with 0.05 Gy (p = 0.000002) and 0.1 Gy (p = 0.000136) led to a significantly lower cell survival in HT29 cells, whereas HT29 cells irradiated with 0.03 Gy did not show a significant decrease in cell survival, when compared to the control group (Table 1).
The adaptive response experiments, i.e. the experiments performed with a pre-irradiation followed by a 2.0 Gy irradiation, did not demonstrate a significant induced radioresistance. The 0.05 Gy pre-irradiation dose even led to a significantly decreased cell survival (p = 0.012249). (Table 2).
GM637 cell studies
An irradiation dose of 0.03 Gy (p = 0.711896) alone did not result in a significantly lower cell survival of GM637 cells, while irradiation doses of 0.05 Gy (p = 0.000003) and 0.1 Gy (p = 0.008301) led to a significantly reduced cell survival (Table 3).
Pre-irradiation doses of 0.03 Gy (p = 0.002591) or 0.1 Gy (p = 0.044575) applied 4 h prior to the 2.0 Gy fraction led to a significantly enhanced cell survival in GM637 cells, when compared to cells irradiated with 2.0 Gy alone. These pre-irradiation doses therefore led to an induced radioresistance in GM637 cells. This effect was most pronounced in the 0.03 Gy experiment. A pre-irradiation of 0.05 Gy led to a slightly increased radioresistance, which was not statistically significant however (p = 0.429477). (Table 4).
HT29 and GM637 cell studies in comparison
An irradiation with 0.03 Gy alone did not have a significant effect on the survival of HT29 and GM637 cells, whereas 0.05 Gy and 0.1 Gy led to a significantly lower cell survival in both cell lines.
The effect of the various pre-irradiation doses applied 4 h prior to the 2.0 Gy fraction varied between HT29 and GM637 cells. Pre-irradiation doses of 0.03 Gy and 0.1 Gy induced a significant radioprotective effect in GM637 fibroblasts, but not in HT29 colorectal carcinoma cells. A pre-irradiation dose of 0.05 Gy led to a significantly lower cell survival in HT29 cells, and a slightly, not significantly, higher survival in GM637 cells. A pre-irradiation with 0.03 Gy seems to therefore protect normal fibroblasts, but not colorectal cancer cells, from radiation-induced damage, while an adaptive dose of 0.05 Gy can lead to a reduced survival of colorectal cancer cells, but not of normal fibroblasts.
Modern radiotherapy uses sophisticated techniques to optimize therapeutic tumour control. Side effects on normal tissues, however, are the single most limiting factor to the therapy. Therefore research in the field of radiation oncology not only needs to focus on maximizing tumour destruction but also on minimizing side effects on normal tissues.
The results of our studies imply that a low-dose pre-irradiation applied 4 h prior to the main irradiation may either cause a reduction of the side effects of radiotherapy of colorectal carcinomas on normal tissues or allow enhanced tumour cell killing while not leading to additional side effects – provided that our findings prove to be true in vivo as well.
In our experiments, 0.03 Gy by itself did not have a significant effect on cell survival, neither on the tumour (HT29) nor on the normal cell line (GM637). When this dose is applied as a pre-irradiation dose it may induce a significant radioprotective effect in GM637 human fibroblasts, but not in HT29 colorectal cancer cells. Provided that this phenomenon not only exists in vitro, but also in vivo, and that our cell model reflects real tissue conditions, and moreover exerts its effect for several dose fractions, reduced side effects may be achieved for radiotherapy of colorectal cancer. Using an adaptive dose to protect normal tissue may allow a dose escalation to result in a destruction of more tumour cells. In that case, a better downstaging may theoretically be achieved thus allowing more radical resections. The addition of a daily dose of 0.03 Gy to the conventional 1.8 Gy or 2.0 Gy fractions would add no more than 0.9 Gy to the total irradiation dose applied in approx. 30 sessions.
Hints for an alternative possible application of pre-irradiation doses in radiotherapy of colorectal cancer might result from our experiments, as well. A pre-irradiation dose of 0.05 Gy led to a significantly lower cell survival in HT29 cells, and a slightly, not significantly, higher survival in GM637 cells. When an adaptive dose of 0.05 Gy can lead to a reduced survival of colorectal cancer cells, but not of normal fibroblasts, the pre-irradiation can help to improve tumour cell killing in cancer therapy while not adding more side effects.
Clinical studies are, however, needed to evaluate whether this assumptions holds true in a clinical setting.
In our study, we have concentrated on the commonly known doses for radio-adaptive response experiments. We have therefore not used a pre-irradiation dose of less than 0.03 Gy. As we demonstrated the pre-irradiation dose of 0.03 Gy to be effective to induce radioresistance in normal fibroblasts, it remains to be investigated, however, whether a dose below 0.03 Gy may also lead to the above mentioned effects.
Lambin et al. demonstrated HT29 cells to be hypersensitive to low radiation doses. While the cell survival response showed a good fit to the linear quadratic model for 2 to 5 Gy, it demonstrated a hyperradiosensitivity for 0.05-0.3 Gy and an induced radioresistance for 0.3-1.0 Gy . This is consistent with our findings that HT29 cells show a significantly lower survival when irradiated with 0.05 Gy or 0.1 Gy alone. A dose of 0.03 Gy was not tested by Lambin et al.. In our experiments 0.03 Gy alone did not have a significant effect on the survival of HT29 cells.
Cell lines known to be relatively radioresistant, e.g. HT29 cells (colorectal cancer) and RT4 cells (bladder cancer), often demonstrate a hyperradiosensitive reaction to low irradiation doses [14, 20, 21]. It remains to be clarified, whether this hyperradiosensitivity is an independent effect or whether it represents the absence of induced radioresistance . The induction of PBP74/mortalin/Grp75, a member of the hsp 70 family, seems to play a role in induced radioresistance in HT29 cells .
A radio-adaptive response can be measured in terms of cell survival – as performed in our study -, of reduction of chromosomal aberrations, of micronuclei formation or of mutations [3, 22–24]. It occurs after pre-irradiation doses of 0.01 Gy  to 1.5 Gy  depending on the cell line examined and on the experimental conditions. In our study, we observed a radio-adaptive response in GM637 cells for 0.03 Gy and 0.1 Gy pre-irradiation doses. The time span of 4 h between pre-irradiation and the challenging dose has been used in the past , however, other intervals have been studied as well with differing results .
The mechanism of the adaptive response is still not completely understood, but it is widely accepted that inducible DNA repair mechanisms play an important role , whereas others believe in decreased damage fixation . Furthermore, stress response, apoptosis pathways, signal cascades, DNA conformation changes, chromosome organization, bystander effects and cell cycle control are probably involved as well [1, 29–33]. Protein synthesis appears to be essential for the induction of an adaptive response  and several genes have been identified that play a crucial role in this phenomenon [6–10]. Recent publications pointed out the role of the MAPKs p38 and ERK1/2 , NF-κB  and activation of Raf and Akt . It is proposed that the radio-adaptive response follows mainly from mutations at the base-sequence level, not the chromosome level,  and involves some components of the nucleotide excision repair pathway . Adaptive and bystander response are presumably linked via reactive oxygen and nitrogen species .
Further experiments will be needed to completely elucidate the mechanism of the adaptive response. In addition, potential clinical applications of the adaptive response need to be studied as well. Our data and previous studies suggest normal and tumour cells to react differently to low pre-irradiation doses. While bladder cancer cells show a hyperradiosensitivity, normal bladder cells demonstrate an induced radioresistance .
In conclusion, we demonstrated a pre-irradiation with 0.03 Gy to protect human fibroblasts (GM637) but not colorectal cancer cells (HT29) from radiation induced damage of a subsequent 2.0 Gy challenging dose and the application of 0.05 Gy prior to the 2.0 Gy fraction to enhance the cell killing of colorectal cancer cells while not additionally damaging normal fibroblasts. If these findings prove to be true in vivo as well this confirms the hypothesis that low pre-irradiation doses may optimize the balance between local tumour control and injury to normal tissue in modern radiotherapy of colorectal cancer, one of the most common neoplasms world-wide.
Mitchell SA, Marino SA, Brenner DJ, Hall EJ: Bystander effect and adaptive response in C3H 10T1/2 cells. Int J Radiat Biol 2004, 80: 465-472. 10.1080/09553000410001725116
Olivieri G, Bodycote J, Wolff S: Adaptive Response of Human Lymphocytes to Low Concentrations of Radioactive Thymidine. Science 1984, 223: 594-597. 10.1126/science.6695170
Azzam EI, Raaphorst GP, Mitchel REJ: Radiation-Induced Adaptive Response for Protection against Micronucleus Formation and Neoplastic Transformation in C3H 10T1/2 Mouse Embryo Cells. Radiat Res 1994, 138: S28-S31. 10.2307/3578755
Vijayalaxmi , Leal BZ, Leal BZ, Deahl TS, Meltz ML: Variability in adaptive response to low dose radiation in human blood lymphocytes: consistent results from chromosome aberrations and micronuclei. Mutat Res 1995, 348: 45-50. 10.1016/0165-7992(95)90020-9
Youngblom JH, Wiencke JK, Wolff S: Inhibition of the adaptive response of human lymphocytes to very low doses of ionizing radiation by the protein synthesis inhibitor cycloheximide. Mutat Res 1989, 227: 257-261. 10.1016/0165-7992(89)90107-3
Sadekova S, Lehnert S, Chow TYK: Induction of PBP74/mortalin/Grp75, a member of the hsp70 family, by low doses of ionizing radiation: a possible role in induced radioresistance. Int J Radiat Biol 1997, 72: 653-660. 10.1080/095530097142807
Amundson SA, Do KT, Fornace AJ: Induction of Stress Genes by Low Doses of Gamma Rays. Radiat Res 1999, 152: 225-231. 10.2307/3580321
Sasaki MS, Ejima Y, Tachibana A, Yamada T, Ishizaki K, Shimizu T, Nomura T: DNA damage response pathway in radioadaptive response. Mutat Res 2002, 504: 101-118.
Coleman MA, Yin E, Peterson LE, Nelson D, Sorensen K, Tucker JD, Wyrobek AJ: Low-Dose Irradiation Alters the Transcript Profiles of Human Lymphoblastoid Cells Including Genes Associated with Cytogenetic Radioadaptive Response. Radiat Res 2005, 164: 369-382. 10.1667/RR3356.1
Cramers P, Atanasova P, Vrolijk H, Darroudi F, van Zeeland AA, Huiskamp R, Mullenders LH, Kleinjans JC: Pre-exposure to Low Doses: Modulation of X-Ray-Induced DNA Damage and Repair? Radiat Res 2005, 164: 383-390. 10.1667/RR3430.1
Okazaki R, Ootsuyama A, Norimura T: TP53 and TP53-Related Genes Associated with Protection from Apoptosis in the Radioadaptive Response. Radiat Res 2007, 167: 51-57. 10.1667/RR0623.1
Sorensen KJ, Attix CM, Chritian AT, Wyrobek AJ, Tucker JD: Adaptive response induction and variation in human lymphoblastoid cell lines. Mutat Res 2002, 519: 15-24.
Andersson HC, Na Chiangmai S: No adaptive response of Chinese hamster ovary cells to low doses of ionizing radiation. Hereditas 1992, 117: 215-222. 10.1111/j.1601-5223.1992.tb00017.x
Schaffer M, Schwarz SB, Kulka U, Busch M, Duehmke E: Adaptive doses of irradiation-an approach to a new therapy concept for bladder cancer? Radiat Environ Biophys 2004, 43: 271-276. 10.1007/s00411-004-0256-5
World Health Organization. Media Centre[http://www.who.int/mediacentre/factsheets/fs297/en/index.html]
Robert Koch Institut. Krebsneuerkrankungen in Deutschland[http://www.rki.de/cln_049/nn_204084/DE/Content/GBE/DachdokKrebs/Datenbankabfragen/Neuerkrankungen/neuerkrankungen__node.html?__nnn=true;http://22.214.171.124/servlet/Trends]
Wouters BG, Skarsgard LD: Low-dose radiation sensitivity and induced radioresistance to cell killing in HT-29 cells is distinct from the adaptive response and cannot be explained by a subpopulation of sensitive cells. Radiat Res 1997, 148: 435-442. 10.2307/3579320
Murray D, Mirzayans R, Scott AL, Allalunis-Turner MJ: Influence of Oxygen on the Radiosensitivity of Human Glioma Cell Lines. Am J Clin Oncol 2003, 26: e169-e177. 10.1097/01.coc.0000091359.11281.f4
Lambin P, Marples B, Fertil B, Malaise EP, Joiner MC: Hypersensitivity of human tumour cell line to very low radiation doses. Int J Radiat Biol 1993, 63: 639-650. 10.1080/09553009314450831
Joiner MC, Lambin P, Malaise EP, Robson T, Arrand JE, Skov KA, Marples B: Hypersensitivity to very-low single radiation doses: Its relationship to the adaptive response and induced radioresistance. Mutat Res 1996, 358: 171-183.
Joiner MC: Induced radioresistance: an overview and historical perspective. Int J Radiat Biol 1994, 65: 79-84. 10.1080/09553009414550111
Ishii K, Misonoh J: Induction of Radio-adaptive Response by Low-dose X-irradiation on Chromosome Aberrations in Human Embryonic Fibroblasts. Physiol Chem Phys & Med NMR 1996, 28: 83-90.
Shadley JD, Wolff S: Very low doses of X-rays can cause human lymphocytes to become less susceptible to ionizing radiation. Mutagenesis 1987, 2: 95-96. 10.1093/mutage/2.2.95
Ueno AM, Vannais DB, Gustafson DL, Wong JC, Waldren CA: A low, adaptive dose of gamma-rays reduced the number and altered the spectrum of S1-mutants in human-hamster hybrid A L cells. Mutat Res 1996, 358: 161-169.
Venkat S, Apte SK, Chaubey RC, Chauhan PS: Radioadaptive Response in Human Lymphocytes in Vitro. JEPTO 2001, 20: 165-175.
Sasaki MS: On the reaction kinetics of the radioadaptive response in cultured mouse cells. Int J Radiat Biol 1995, 68: 281-291. 10.1080/09553009514551211
Skov KA: Molecular, Cellular, and Genetic Basis of Radiosensitivity at Low Doses: A Case of Inducible Repair? Radiat Res 1994, 138: S1-S4. 10.2307/3578840
Szumiel I: Adaptive response: stimulated DNA repair or decreased damage fixation? Int J Radiat Biol 2005, 81: 233-241. 10.1080/09553000500077047
Boothman DA, Meyers M, Fukunaga N, Lee SW: Isolation of x-ray-inducible transcripts from radioresistant human melanoma cells. Proc Natl Acad Sci USA 1993, 90: 7200-7204. 10.1073/pnas.90.15.7200
Parkhomenko IM, Periscvili G, Turovetskii V, Raev B, Lubov D, Brovko Y, Kudrjascov Y, Rubin A: Molecular-cellular Mechanisms of Low Doses Biological Action: Effects on the Systems of Cell Regulation and the Ways of the Low Doses Early Detection. In Low Dose irradiation and biological defence mechanisms. Edited by: Sugahara T, Sagan LA, Aoyama T. Amsterdam London New York Tokyo: Excerpta Medica, Elsevier Science Publishers B.V; 1992:423-425.
Belyaev IY, Spivak IM, Kolman A, Harms-Ringdahl M: Relationship between radiation induced adaptive response in human fibroblasts and changes in chromatin conformation. Mutat Res 1996, 358: 223-230.
Schwartz JL: The Role of Constitutive and Inducible Processes in the Response of Human Squamous Cell Carcinoma Cell Lines to Ionizing Radiation. Radiat Res 1994, 138: S37-S39. 10.2307/3578757
Nuta O, Darroudi F: The impact of the bystander effect on the low-dose hypersensitivity phenomenon. Radiat Environ Biophys 2007. 10.1007/s00411-007-0145-9
Kim CS, Kim JM, Nam SY, Yang KH, Jeong M, Kim HS, Lim YK, Kim CS, Jin YW, Kim J: Low-dose of Ionizing Radiation Enhances Cell Proliferation Via Transient ERK1/2 and p38 Activation in Normal Human Lung Fibroblasts. J Radiat Res 2007, 48: 407-415. 10.1269/jrr.07032
Ahmed KM, Li JJ: NF-κB-mediated adaptive resistance to ionizing radiation. Free Radic Biol Med 2008, 44: 1-13. 10.1016/j.freeradbiomed.2007.09.022
Kim CS, Kim JK, Nam SY, Yang KH, Jeong M, Kim HS, Kim CS, Jin YW, Kim J: Low-Dose Radiation Stimulates the Proliferation of Normal Human Lung Fibroblasts Via a transient Activation of Raf and Akt. Mol Cells 2007, 24: 424-430.
Yatagai F, Umebayashi Y, Masamitsu H, Sugasawa K, Takayama Y, Hanaoka F: Mutagenic radioadaptation in human lymphoblastoid cell line. Mutat Res 2008, 638: 48-55.
Hafer K, Iwamoto KK, Scuric Z, Schiestl RH: Adaptive Response to Gamma Radiation in Mammalian Cells Proficient and Deficient in Components of Nucleotide Excision Repair. Radiat Res 2007, 168: 168-174. 10.1667/RR0717.1
Matsumoto H, Hamada N, Takahashi A, Kobayashi Y, Ohnishi T: Vanguards of Paradigm Shift in Radiation Biology: Radiation-Induced Adaptive and Bystander Responses. J Radiat Res 2007, 48: 97-106. 10.1269/jrr.06090
The authors declare that they have no competing interests.
SBS: designed protocol, conducted data evaluation, wrote the article
PMS: designed protocol, conducted data evaluation, wrote the article
UK: collected data, statistical analysis, laboratory controlling
BEW: statistical analysis, critical review of the manuscript
RH: biological technical assistant
MS: designed protocol, conducted data evaluation, critical review of the manuscript, group supervisor
All authors read and approved the final manuscript.
Silke B Schwarz, Pamela M Schaffer contributed equally to this work.