Comments for:

---

Response to letter from Leffert, Maeda and Karin

Submitter: Tom Luedde | Tom.Luedde@web.de

Dept. of Gastroenterology, Hepatology & Endocrinology, Hannover Medical School, Hannover, Germany

Published September 6, 2005

We appreciate the interest of Leffert, Maeda and Karin in our study, who, with their previous work, have contributed significantly towards the understanding of the function of the NF-κB signaling pathway in the liver. In their letter, Leffert et al. refer to the differences between their own experiments published earlier (1) and our findings reported in the current study (2).

For generating a liver-specific conditional knockout of IKK2/IKKβ, two different loxP-flanked IKK2/IKKβ alleles and two different liver- specific Cre transgenic lines were used in the two studies. Maeda et al used a conditional Ikk2 allele where exon 3 is flanked by loxP sites (1), while in our study we used a conditional IKK2 allele where exons 6 and 7 are flanked by loxP sites. Maeda et al. used the Alb-Cre mouse (3) which deletes floxed alleles in hepatocytes perinatally (the conditional knockout mouse is referred to as IKKβ^Δhep). In our study, we used the Alfp-cre mouse (4), which utilizes an AFP-enhancer element together with the albumin promoter driving Cre-recombinase expression in the liver from embryonic life. We termed Alfp-Cre/Ikk2FL/FL mice: IKK2Δhepa.

Lettert et al. expressed the following concerns about our experiments:
1) Leffert et al. disagree with our conclusion “that targeted deletion of hepatocyte IKKbeta does not sensitize hepatocytes to TNFalpha-induced apoptosis”.

In our study, we injected recombinant TNF into mice, which had little or no effect in mice lacking IKK2/IKKβ in the liver, whereas this treatment led to massive liver failure and hepatocyte apoptosis in mice lacking the regulatory subunit NEMO/IKKγ. Our title is based on this clear in vivo finding, which demonstrates that complete inhibition of NF-κB (by NEMO deletion) sensitizes hepatocytes to TNF induced death, while deletion of IKK2 does not lead to sufficient inhibition of NF-κB activation and therefore does not sensitize these cells to TNF toxicity. Maeda et al did not inject recombinant TNF into mice but used LPS and ConA, which in addition to inducing secretion of TNF also activates many pathways in multiple cell-types leading to the expression of various cytokines other than TNF. Based on their finding that ConA, but not LPS, induces liver damage in the IKKβ ^Δhep mice, they claim that transmembrane TNF (induced by ConA treatment) kills IKK2-deficient cells while soluble TNF (induced by LPS) does not have such an effect. This conclusion does not take into account that the differences could also be attributed to the fact that a signal induced by LPS but not by ConA, either directly or indirectly via an LPS-inducible mediator, to hepatocytes might activate a protective mechanism against TNF killing. In fact, in figure 6D of their manuscript, the authors demonstrate that stimulation of primary hepatocytes from IKKβ^Δhepmice with soluble mouse TNF led to a significant increase in hepatocyte death compared to wildtype cells (1), which does not support the statement made in the title of their manuscript. In both studies, LPS injection in vivo did not result in significantly increased damage in hepatocytes lacking IKK2/ IKKβ. In our study, injection of ConA did not result in a significant difference of hepatocyte damage between IKK2Δhepa mice and controls (2), whereas in the study by Maeda et al, the authors describe a stronger liver damage in their IKKβ^Δhep mice (1) upon ConA injection. This difference might be explained by a stronger inhibition of NF- κB activation in their mice, as suggested in our discussion. Finally, Maeda et al. claim, based on their data, that “inhibition of NF-κB activity is not sufficient for induction of liver failure even in the presence of very high levels of circulating TNFα”. This statement strongly contrasts numerous studies that were published before (5-9) and lacks a molecular explanation. Furthermore, in our study we show a strong correlation between NF-κB activation and the protection from TNF- induced apoptosis and liver failure.

2) They “expect that IKK2Δhepa mice undergo significant embryonic lethality, with targeted survivors generated in a non-Mendelian fashion” .

This is not the case. During the course of our study a large number of IKK2Δhepa mice and their respective control littermates were analyzed, which came from breeding pairs where both parents carried two loxP-flanked Ikk2 alleles and one parent expressed the Alfp-Cre transgene (Ikk2FL/FL x Alfp-Cre/Ikk2FL/FL). In fact, from more than 100 mice generated by this breeding strategy, 51% were cre-positive and 49% cre-negative.

3) They “suggest that precocious IKKbeta deletion occurs easily in IKK2Δhepa mice giving rise to significant embryonic lethality. Possibly, in their system, the IKK2Δhepa mice who survive upregulate IKKα or undergo other molecular changes and these other changes – not targeted IKKbeta-deletion -- may account for the results they report.”

As mentioned above, no significant embryonic lethality of IKK2Δ hepa mice was observed. Moreover, NF-κB activation upon TNF stimulation was examined in IKK2Δmice (figure 3A of our publication) in addition to IKK2Δhepa mice (2). As stated in the materials and methods section, IKK2 deletion in IKK2Δmice was achieved through induction of 6 week-old IKK2^Fl/Fl-Mx-cre mice with poly-deoxyinosine- deoxycytodine (poly I/C). This led to a complete knockout of IKK2 in the liver, which is demonstrated in figure 1G/H. Although in this mouse IKK2 is deleted during adulthood (later than both the IKK2Δhepa and the IKKβ ^Δhep mice), the mice show the same phenotype in terms of TNF-dependent NF-κB activation in the liver as the IKK2Δhepa mice (2). In hepatocytes isolated from IKK2Δ mice NF-κB is activated upon TNF-stimulation, in hepatocytes isolated from NemoΔ mice NF-κB is not activated (figure 3E). Both IKK2Δ and NemoΔ mice had the respective IKK subunit deleted through poly I/C induction during adulthood. Our published results therefore refute the alternative explanations offered by Leffert et al.

4) Leffert et al. claim that we show “hardly any biochemical characterization of the effect of the mutation on the composition of the IKK complex, …”

In figure 4 of our study, we performed IP-studies in order to provide data to characterize the effect of the mutation on the composition of the IKK complex. In figure 4B, we show that levels of IKK1 are not upregulated in IKK2Δhepa mice. Moreover, in figure 4A not more but less IKK1 is bound to the NEMO subunit (2), suggesting that upregulation of IKK1 does not represent a compensatory mechanism in IKK2Δhepa mice.

In summary, we would like to emphasize that we appreciate the interesting comments of Leffert et al. on our study. We would be happy to try to clarify the differences between the two published conditional IKK2 knockout mice in a collaborative effort. However, we are surprised by the demand from Leffert et al., that a “burden of proof” should have been placed on us given their earlier publication. It is difficult to understand what additional proof we should provide other than clear experimental data. It is our belief that conclusions based on solid results that are obtained through well-controlled experiments should be published, even if they do not agree with previously reported studies. After all, we find the differences between our study and that of Maeda et al extremely interesting and we believe that by looking more closely into this apparent contradiction we will gain further important insights into the function of IKK subunits and NF-κB in the liver.

References

1. Maeda,S., Chang,L., Li,Z.W., Luo,J.L., Leffert,H., and Karin,M. 2003. IKKbeta is required for prevention of apoptosis mediated by cell-bound but not by circulating TNFalpha. Immunity. 19:725-737.

2. Luedde,T., Assmus,U., Wustefeld,T., Meyer,z., V, Roskams,T., Schmidt-Supprian,M., Rajewsky,K., Brenner,D.A., Manns,M.P., Pasparakis,M. et al. 2005. Deletion of IKK2 in hepatocytes does not sensitize these cells to TNF-induced apoptosis but protects from ischemia/reperfusion injury. J.Clin.Invest 115:849-859.

3. Postic,C., Shiota,M., Niswender,K.D., Jetton,T.L., Chen,Y., Moates,J.M., Shelton,K.D., Lindner,J., Cherrington,A.D., and Magnuson,M.A. 1999. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J.Biol.Chem. 274:305-315.

4. Kellendonk,C., Opherk,C., Anlag,K., Schutz,G., and Tronche,F. 2000. Hepatocyte-specific expression of Cre recombinase. Genesis. 26:151 -153.

5. Beg,A.A. and Baltimore,D. 1996. An essential role for NF-kappaB in preventing TNF-alpha-induced cell death. Science 274:782-784.

6. Bellas,R.E., FitzGerald,M.J., Fausto,N., and Sonenshein,G.E. 1997. Inhibition of NF-kappa B activity induces apoptosis in murine hepatocytes. Am.J.Pathol. 151:891-896.

7. Doi,T.S., Marino,M.W., Takahashi,T., Yoshida,T., Sakakura,T., Old,L.J., and Obata,Y. 1999. Absence of tumor necrosis factor rescues RelA- deficient mice from embryonic lethality. Proc.Natl.Acad.Sci.U.S.A 96:2994 -2999.

8. Grossmann,M., Metcalf,D., Merryfull,J., Beg,A., Baltimore,D., and Gerondakis,S. 1999. The combined absence of the transcription factors Rel and RelA leads to multiple hemopoietic cell defects. Proc.Natl.Acad.Sci.U.S.A 96:11848-11853.

9. Nagaki,M., Naiki,T., Brenner,D.A., Osawa,Y., Imose,M., Hayashi,H., Banno,Y., Nakashima,S., and Moriwaki,H. 2000. Tumor necrosis factor alpha prevents tumor necrosis factor receptor-mediated mouse hepatocyte apoptosis, but not fas-mediated apoptosis: role of nuclear factor-kappaB. Hepatology 32:1272-1279.

Sincerely,
Tom Luedde and Manolis Pasparakis
EMBL Mouse Biology Program, Via Ramarini 32, 00016 Monterotondo, Italy
Christian Trautwein
Medizinische Klinik III, UKA, RWTH Aachen, Pauwelsstrasse 30, 52074 Aachen, Germany

---

letter about Luedde et al

Submitter: Hyam Leffert | hleffert@ucsd.edu

University of California, San Diego

Published August 24, 2005

We wish to take issue with the conclusion of T. Luedde et al that targeted deletion of hepatocyte IKKβ does not sensitize hepatocytes to TNFα-induced apoptosis.¹ On the contrary, our prior study showed unequivocally that hepatocyte IKKβ is required for prevention of apoptosis mediated by cell-bound but not by circulating TNFα (ref. 2).
To circumvent the problem of embryonic lethality and enable investigation of the role of IKKβ in liver function and hepatocarcinogenesis³, we generated the conditional knockout mouse strain called Ikkβ^Δhep in two steps. First, we created a new strain of normal mice called Ikkβ^F/F. These animals carry an homozygous 'floxed' IKKβ allele in which bacteriophage loxP recombination targets of Cre recombinase flank the catalytic kinase domain in exon 3 of the Ikkβ gene. In the second step, Ikkβ^F/F mice were crossed with Alb-Cre transgenic mice⁴ to generate Ikkβ^Δhep mice. In crosses of this kind, Cre-mediated excision of floxed alleles, driven by the hepatocyte-specific albumin enhancer-promoter, is delayed until perinatal life^2-5; and, Ikkβ^Δhepprogeny homozygous for both alleles undergo normal embryonic development, growth and lifespan^2,3.
In contrast, Luedde et al reported on studies with a different hepatocyte-specific IKKβ-knock-out mouse Ikk2Δhepa) generated with floxed exons 6-to-7-Ikk2^f/f and Alfp-Cre strains¹. Owing to early developmental activation of the Alfp-cre construct, the Ikk2 deletion is created embryonically early in prenatal life. Under these conditions, we would expect that Ikk2 Δhepa mice undergo significant embryonic lethality, with targeted survivors generated in a non-Mendelian fashion.
In their Discussion, Luedde et al. suggested that the deleted IKKβ allele in Ikkβ^Δhep mice generates a dominant negative mutation¹. This is extremely unlikely because the Ikkβ^Δhep deletion generates a frame shift that destabilizes the mRNA in these mice. Therefore, IKKβ is a true null mutation in Ikkβ^Δhep mice. In addition, in Ikkβ^Δhep-deleted hepatocytes, IKKα is easily immunoprecipitated with IKKγ (NEMO)², whereas, in contrast to statements in Luedde et al, a dominant negative mutation would disrupt such interaction. Therefore we believe that Luedde et al's explanation for the different results between the different targeted Ikkβ knockout strains is quite unlikely. Instead, we suggest that precocious IKKβ deletion occurs easily in Ikk2 Δhepa mice giving rise to significant embryonic lethality. Possibly, in their system, the Ikk2 Δhepa mice who survive upregulate IKKα or undergo other molecular changes and these other changes – not targeted IKKβ-deletion -- may account for the results they report.
We think, given our earlier publication on this topic, that the burden of proof should have been placed on Luedde et al. Yet there is hardly any biochemical characterization of the effect of the mutation on the composition of the IKK complex, nor mention of the fact that many of their mutant mice might die in utero or evidence provided to the contrary.
¹Luedde T, Assmus U, Wustefeld T, Meyer zu Vilsendorf A, Roskams T, Schmidt-Supprian M, Rajewsky K, Brenner DA, Manns MP, Pasparakis M, Trautwein C. Deletion of IKK2 in hepatocytes does not sensitize these cells to TNF-induced apoptosis but protects from ischemia/reperfusion injury. J Clin Invest. 115:849-859, 2005.
²Maeda S, Chang L, Li Z-W, Luo J-L, Leffert H, Karin M. IKKβ is required for prevention of apoptosis mediated by cell-bound but not by circulating TNFα. Immunity 19:725-737, 2003.
³Maeda S, Kamata H, Luo J-L, Leffert H, Karin M. IKKβ couples hepatocyte death to cytokine-driven compensatory proliferation that promotes chemical hepatocarcinogenesis. Cell 121: 977-990, 2005.
⁴Postic C, Shiota M, Niswender KD, Jetton TL, Chen Y, Moates JM, Shelton KD, Lindner J, Cherrington AD, Magnuson MA. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J Biol Chem. 274:305-315, 1999.
⁵Postic C, Magnuson MA. DNA excision in liver by an albumin-Cre transgene occurs progressively with age. Genesis 26:149-150, 2000.
Sincerely,
Shin Maeda Division of Gastroenterology
The Institute for Adult Diseases Asahi Life Foundation
1-6-1 Marunouchi, Chiyoda-ku Tokyo 100-0005
Hyam Leffert and Michael Karin
Department of Pharmacology
University of California at San Diego La Jolla CA 92093