The Journal of Neurochemistry
Matters Arising

---

---

Letter to the Editor

Dear Sir,
In the recent issue of your journal, Greene et al provided an excellent, informative review on the antiepileptic mechanisms of ketone bodies. The authors propose that epilepsy ultimately involves and results from altered brain energy homeostasis. They briefly discuss glucose uptake into brain, but unfortunately did not address the GLUT1 deficiency syndrome, a novel condition that precisely supports this message.

GLUT1 deficiency syndrome (OMIM 606777) is caused by a defect in the facilitative glucose transporter GLUT1 at the blood-brain barrier and in brain cells, resulting in seizures, developmental delay, and a complex motor disorder in early childhood. The laboratory hallmark of this condition is unexplained hypoglycorrhachia. The GLUT1 defect is confirmed by impaired glucose uptake into erythrocytes and heterozygous mutations in the GLUT1-gene (1p35-31.3), suggesting an autosomal dominant disease resulting from haploinsufficiency. In these patients, introducing a ketogenic diet is a very effective treatment as ketones enter the brain by the MCT1 transporter and serve as an alternative fuel for brain energy metabolism. In the vast majority of these patients an effective seizure control is achieved within days of introducing the diet and anticonvulsants can be discontinued.

Without doubt the anticonvulsant mechanisms of ketone bodies in the majority of children with intractable epilepsy are complex as skillfully discussed by the authors. However, I feel it would have been informative to bring the GLUT1 deficiency syndrome to the attention of the reader. The conditon offers an additional, very compelling anticonvulsant mechanism of ketone bodies and supports the authors' conclusion that elevated ketone bodies can replace lost energy from glucose - at least in this form of epilepsy.

Sincerely yours,
Jörg Klepper, MD

References:

1. De Vivo DC, Trifiletti RR, Jacobson RI, Ronen GM, Behmand RA, Harik SI (1991) Defective glucose transport across the blood-brain barrier as a cause of persistent hypoglycorrhachia, seizures, and developmental delay. N Engl J Med 325:703-709.

2. Seidner G, Garcia-Alvarez M, Yeh JI, O'Driscoll K, Klepper J, R, Stump TS, Wang D, Spinner NB, Birnbaum MJ, De Vivo DC (1998) Glut-1 deficiency syndrome caused by haploinsufficiency of the blood-brain barrier hexose carrier. Nature Gen 18:1-4

3. Klepper J, Voit T. (2002) Facilitated glucose transporter protein type 1 (GLUT1) deficiency syndrome: impaired glucose transport into brain - a review. Eur J Pediatr 161:295-304

---

Corrigendum: Functional and molecular characterization of individual olfactory neurons. by, Johannes Noé and Heinz Breer, Journal of Neurochemistry 1998, vol. 7, No. 6, pp. 2286-2293

We have to extend the previous corrigendum by slightly changing the text and by including the original and modified version of figures 6 and 8.

---

Dear Sir:

I read with interest the article by:
Gozal E., Gozal D., Pierce W. M., Thongboonkerd V., Scherzer J. A., Sachleben L. R., Jr, Brittian K. R., Guo S.-Z., Cai J. and Klein J. B. (2002). Proteomic analysis of CA1 and CA3 regions of rat hippocampus and differential susceptibility to intermittent hypoxia. J. Neurochem. 83: 331-345.

It is always a puzzle how CA1 and CA3 hippocampal domains differentially respond to CNS injury; in this case intermittent hypoxia. Proteomics is a novel approach and the authors did a commendable job. However it would be helpful if the authors could clarify the following issues.

1. The authors mentioned that they identified 99 proteins (Table 1 of the original publication). However, there seems to be some duplication based on their description, as some of the proteins have the same Swiss Prot # and MW. Some proteins which are obviously different have the same Swiss Prot #. These are listed in the table on the next page.

2. Western blot of hsp70 (Fig. 2a) shows much greater expression in CA3 compared to CA1 under normoxic conditions. However, Table 2 shows that HSP70 (fragment) is expressed 2.8-fold greater in CA1. There was no explanation of this apparent discrepancy. What is the relationship between hsp70 and HSP70 (fragment)? Is this a 42kD fragment? This is an important issue since Table 3b shows that expression of HSP70 (fragment) decreased about 50% following IH.

3. According to Table 2, TOAD 64 is expressed more in CA3 than CA1. This does not agree with the immunostaining in Fig. 3. The 40x and 100x magnifications highlight the CA1 region. In the 5x magnification, the CA3 does not seem to express TOAD 64.

4. The results states that all the metabolism related proteins were expressed more in CA1 region than in CA3. This is not strictly true. The discussion states that fructose bisphosphate aldolase C1 (FBA) was the only metabolic-related protein that displayed enhanced expression in the CA3 region.

5. In Table 3a, is Vacuolar ATP synthase subunit B different from Vacuolar ATP synthase, B subunit?

6. Table 3b. Ratios of mean densitometry values of CA1 protein in IH/CA3 protein in normoxia seem inappropriate. Is this ratio of CA3 protein in IH/CA3 protein in normoxia?

Dr. Rao M. Adibhatla
Senior Scientist
H4-330; CSC; Department of Neurological Surgery
600 Highland Avenue; Univ. of Wisconsin; Madison
Phone: (608) 263-1791; Fax: (608) 263-1409; email: adibhatl@neurosurg.wisc.edu

Observations made from Table 1 of Gozal et al J Neurochem. 83, 331-345 (2002). Posted January 2, 2003

ID	Protein/alternate name or description	pI	MW	Swiss Prot #	Comment
1	Neurofilament Triplet L, big	4.63	61355	P19527	Appear to be duplicates.
13	Neurofilament Triplet L	4.63	61355	P19527
17	14-3-3 delta	4.73	27925	P35215	Appear to be duplicates.
18	14-3-3 zetz/delta	4.73	27925	P35215
32	Antioxidant protein 2	6	24924	O35244	Appear to be duplicates.
33	antioxidant protein 2 basic	6	24924	O35244
38	alpha-enolase basic	6.5	46984	P04764	Appear to be duplicates.
39	alpha-enolase	6.5	46984	P04764
41	HSP 60	5.35	58061	P19226	Appear to be duplicates.
67	HSP 60 protein	5.35	58061	P19226
43	ATP synthase α chain	8.28	55537	P15999	Appear to be duplicates.
76	ATP synthase H + transporting α subunit	8.28	55537	P15999
69	Vacuolar ATP synthase subunit B	5.57	56875	P50517	Appear to be duplicates.
85	Vacuolar ATP synthase, B subunit	5.57	56875	P50517
2	Vacuolar ATPase subunit B	5.7	56550	P50517	Same Swiss Prot # as 69 and 85
78	Phosphoglycerate kinase 1	7.52	44907	P09411	Appear to be duplicates.
83	Phosphoglycerate kinase	7.52	44907	P09411
58	Vacuolar ATPase catalytic subunit A	5.62	68268	P50516	Appear to be duplicates
73	Vacuolar ATP synthase catalytic subunit A ubiquitous form	5.62	68268	P50516	pI, MW and SP# were shifted in the Table 1
61	UBF transcription factor short form	5.67	89460	Q60459	Appear to be duplicates based on identical MW.
74	Ribosomal transcription factor UBF1	-	89460	-
3	TOAD 64	6.3	62277	P47942	Appear to be duplicates.
42	DRP2, TOAD	6.3	62277	P47942
79	DRP-2 (TOAD)	7.52	62638	P09411	Appears to be different, but alternate name is same as ID #42
4	ß-tubulin big, T-beta 15	4.8	49963	P04691	Same Swiss Prot# with slight difference in MW
11	ß tubulin, T-beta 15	4.8	49938	P04691
7	ß actin	5.29	42068	P02570	Same Swiss Prot# with slight difference in MW
59	cytoplasmic ß actin 1	5.4	41750	P02570
62	HSP 70 Big	-	42031	-	Are these different from the hsp70 shown in Fig. 2a (~70kD)?
46	HSP 70	-	42159	-
--	HSP 70 (fragment)	-	-	-	Shown in Table 3B, but not included in Table 1.
56	SRP 14K	-	9380	O55098	Have the same Swiss Prot# but differs in MW
57	Serine Threonine Kinase 10	6.53	57583	O55098
48	Hemoglobin	-	-	-	Appears to be unidentified
52	Myosin heavy chain?	-	-	-	Appears to be unidentified
97	Myosin heavy chain α isoform	5.59	224287	P02563	MW larger than 150K; however detection limit as stated on p. 338 is 150K.

 

Reply from the authors:

1. The authors mentioned that they identified 99 proteins (Table 1 of the original publication).However, there seems to be some duplication based on their description, as some of the proteins have the same Swiss Prot # and MW. Some proteins which are obviously different have the same Swiss Prot #. These are listed in the table on the next page.

Dr. Adibhatla correctly notes that our protein database identifies 99 protein spot forms on the gels, not 99 different proteins. The overlap results from multiple forms of the same protein that result from post-translational modification, cleavage products or isoforms. Instances where different protein names are given to the same protein result from the lack of uniform nomenclature in proteomic databases used in peptide mass fingerprinting.
Two typographical errors were made in Table I. The SwissProt entry for protein number 72 should be P49803, not P19803. The entry for number 79 should be P47942.

2. Western blot of hsp70 (Fig. 2a) shows much greater expression in CA3 compared to CA1 under normoxic conditions. However, Table 2 shows that HSP70 (fragment) is expressed 2.8-fold greater in CA1. There was no explanation of this apparent discrepancy. What is the relationship between
hsp70 and HSP70 (fragment)? Is this a 42kD fragment? This is an important issue since Table 3bshows that expression of HSP70 (fragment) decreased about 50% following IH.

Dr. Adibhatla correctly notes that we identified two forms of Hsp70; the native Hsp70 (labeled Hsp70 'big") is expressed on the 2D gels at ca. 70kDa. Table I incorrectly lists the molecular size of Hsp70 "big" as 42 kDa.
We have further analyzed the mass spectra of the Hsp70 "small" and find that it is the 44 kDa N terminus domain of Hsp70 that possesses ATPase activity.

3. . According to Table 2, TOAD 64 is expressed more in CA3 than CA1. This does not agree with the immunostaining in Fig. 3. The 40x and 100x magnifications highlight the CA1 region. In the 5x magnification, the CA3 does not seem to express TOAD 64.

TOAD64 expression was extremely variable in the immunostained sections. The goal of the immunohistochemistry experiments was simply to show that TOAD64 was expressed in the hippocampus as it had only been observed previously in the dentate gyrus. The quantification from 2D gels represented data from seven separate experiments and we feel this is more reliable.

4. The results states that all the metabolism related proteins were expressed more in CA1 region than in CA3. This is not strictly true. The discussion states that fructose bisphosphate aldolase C1 (FBA) was the only metabolic-related protein that displayed enhanced expression in the CA3 region.

Dr. Adibhatla incorrectly interprets our observations regarding the expression of metabolism related proteins in CA1 and CA3. In the Results section of our paper we state "All metabolism-related proteins were enzymes involved in cellular energy metabolism, and showed higher expression levels in the CA1 than in the CA3." It was our intent in this sentence to indicate that all the metabolism-related proteins were involved in cellular energy metabolism, not that all metabolism-related proteins were preferentially expressed in CA1 or CA3. We regret any confusion that the sentence structure caused.

5. In Table 3a, is Vacuolar ATP synthase subunit B different from Vacuolar ATP synthase, B subunit?

Dr. Adibhatla is correct in noting that the ATP synthase subunit B is identical to vacuolar ATP synthase B subunit and as in question one represents a different spot form on the 2D gel. The lack of uniformity in protein databases is the cause of the different nomenclature.

6. Table 3b. Ratios of mean densitometry values of CA1 protein in IH/CA3 protein in normoxia seem inappropriate. Is this ratio of CA3 protein in IH/CA3 protein in normoxia?

Dr. Adibhatla is correct and a typo has occurred in the legend of table 3b which should have read:"Ratios of mean densitometry values of CA3 proteins…"

---

 

On the possible spontaneous generation of free radicals by Alzheimer Aß peptides (posted November 18, 2001)

To the Editors:

In a recent publication, Monji et al. (2001) reported the results of an investigation designed to establish whether or not the Aß 1-40 and Aß 1-42 amyloid peptides, implicated in Alzheimer's disease, spontaneously generate free radicals, and in particular to correlate free radical formation with the aggregation state of the peptides. Electron spin resonance (ESR) spectroscopy was employed, in conjunction with the spin-trapping technique, to monitor any free radicals that might be generated. The peptides were incubated, at 37oC for up to 72 hours, either in phosphate buffered saline (PBS) or phosphate buffer, at pH 7.4, in the presence of N-tert-butyl-a-phenylnitrone (PBN) as the spin-trap. Very weak 4-line ESR spectra, with a(N) 1.45 and a(Hb) 1.45 mT, were observed during the incubation of Aß 1-40 or Aß 1-42 in PBS. We agree that the observed spectrum has been correctly assigned to tert-butylhydroaminoxyl rather than an adduct of PBN itself.

The authors have interpreted this observation as evidence for the spontaneous generation of peptide-derived radicals during Aß incubation, which were then trapped by PBN. They suggest that the tert-butylhydroaminoxyl radical is formed following the breakdown of the PBN-(1-40)-peptidyl and PBN-(1-42)-peptidyl adducts and that the observed spectra are, therefore, evidence for the spontaneous formation of radicals from these two peptides. This interpretation originates from similar experiments performed previously by Butterfield and co-workers (Vardarajan et al. 2000). However, we believe that there is a highly plausible alternative explanation for the origin of the 4-line spectrum that does not invoke the concept of spontaneously generated peptide radicals.

Dikalov et al. (1999) have shown that the weak spectrum of tert- butylhydroaminoxyl can arise following the metal-catalysed auto-oxidation of N-tert-butylhydroxylamine which can be present as an impurity in some samples of PBN. Clearly, PBN needs to be carefully purified before undertaking experiments such as these. However, Butterfield and co-workers (Vardarajan et al. 2000) have shown that the spectrum of tert-butylhydroaminoxyl is still generated slowly in the presence of Aß1-40 or Aß1-42 even when using rigorously purified PBN. The results for Aß 1-40, employing high purity PBN, have been confirmed in our own laboratories (Turnbull et al. 2001a). Importantly, Butterfield and ourselves did not observe any spectra in the absence of the peptide, and so we are forced to accept the conclusion that the presence of Aß is required for the formation of tert-butylhydroaminoxyl. The key question is, how does this occur? Of particular relevance are some recent reports by Bush and co-workers that Ab 1-40 and Aß 1-42 can generate hydrogen peroxide through metal ion reduction (Huang et al. 1999). The potential formation of hydrogen peroxide is intriguing. It is a strong oxidant and, if formed in the presence of redox-active transition metal ions such as Fe and Cu, would result in the formation of the hydroxyl radical via Fenton's reaction. In this respect, it is important to note that not only does chelexation fail to totally remove Cu and Fe from buffered solutions (Turnbull et al. 2001b; Jobling et al. 2001) but, also, that Aß peptides themselves contain significant amounts of these (and other) metal ions (Turnbull et al. 2001b). There is no doubt, of course, that any hydroxyl radicals formed in this way would be trapped by PBN but, in aqueous solution, the resulting adduct is unstable and spontaneously transforms to tert-butylhydroaminoxyl (Kotake and Janzen 1991). In addition, PBN itself is prone to oxidation and hydrolysis to tert-butylhydroaminoxyl and this process would be enhanced in the presence of hydrogen peroxide. Consequently, we believe that the very weak spectrum of tert-butylhydroaminoxyl observed during the incubation of Aß fortuitously monitors hydrogen peroxide and hydroxyl radical production rather than the spontaneous formation of peptide-derived radicals. Whilst we do not necessarily disagree with the conclusions of Monji et al. (2001) regarding the stage of Aß peptide aggregation during which ESR spectra can be observed, we feel that a great deal more care should be exercised in interpreting the origins of these spectra. Spin-trapping experiments can suffer from a number of potential problems. A particular adduct may have more than one source, radicals present in comparatively high concentration may not be trapped efficiently, and some adducts can have limited stability.

We are further concerned that Monji et al. (2001) claim that the 4-line ESR spectrum observed in the presence of Aß was "completely abolished" when deferroxamine or catalase were added as controls, when a careful inspection of their illustrated spectra indicates that this is not so. More importantly, a small residual spectrum is apparent in the absence of the Aß peptide which might well indicate the presence of impurities in the sample of PBN employed.

Despite the concerns expressed above, we support the hypothesis that the direct formation of hydrogen peroxide or hydroxyl radicals from Aß may be crucial in explaining its cytotoxic properties. This now also appears to be true for the a-synuclein protein implicated in Parkinson's disease (Turnbull et al. 2001). However, we are not convinced that peptidyl free radicals are of primary importance in the toxic mechanism, since all of the published ESR data, including those of Monji et al. (2001), can be explained via the formation of hydrogen peroxide. Finally, it should be born in mind that the hydrogen peroxide, or any hydroxyl radicals formed from it, could subsequently attack the peptide to produce peptidyl radicals. In this case, various peptidyl radicals might still be detectable. However, it is more likely to be the hydrogen peroxide or hydroxyl radicals that are directly responsible for the Aß-induced cell death.

*Brian J. Tabner
Stuart Turnbull
Omar M. A. El-Agnaf
David Allsop
Department of Biological Sciences and *Department of Environmental Sciences,
Lancaster University,
Lancaster LA1 4YQ, UK

References

Dikalov, S. I., Vitek, M. P., Maples, K. R. and Mason, R. P. (1999) Amyloid b peptides do not form peptide-derived free radicals spontaneously, but can enhance metal-catalysed oxidation of hydroxylamines to nitroxides. J. Biol. Chem. 274, 9392-9399.

Huang, X., Atwood, C. S., Hartshorn, M. A., Multhaup, G., Goldstein, L. E., Scarpa, R. C., Cuajungco, M. P., Gray, D. N., Lim, J., Moir, R D., Tanzi, R. E. and Bush, A. I. (1999) The Ab peptide of Alzheimer's disease directly produces hydrogen peroxide through metal ion reduction. Biochemistry 38, 7609-7616.

Jobling, M. F., Huang, X., Stewart, L. R., Barnham, K. J., Curtain, C., Volitakis, I., Perugini, M., White, A. R., Cherny, R. A., Masters, C. L., Barrow, C. J., Collins, S. J., Bush, A. I. and Cappai, R. (2001) Copper and zinc binding modulates the aggregation and neurotoxic properties of the prion peptide PrP106-126. Biochemistry 40, 8073-8084.

Kotake, Y. and Janzen, E. G. (1991) Decay and fate of the hydroxyl radical adduct of a-phenyl-N-tert-butylnitrone in aqueous media. J. Am. Chem. Soc. 113, 9503-9506.

Monji, A., Utsumi, H., Ueda, T., Imoto, T., Yoshida, I., Hashioka, S., Tashiro, K-i. and Tashiro, N. (2001) The relationship between the aggregational state of the amyloid-b peptides and free radical generation by the peptides. J. Neurochem. 77, 1425-1432.

Turnbull, S., Tabner, B. J., El-Agnaf, O. M. A., Twyman, L. J. and Allsop, D. (2001a) New evidence that the Alzheimer b-Amyloid peptide does not spontaneously form free radicals: An ESR study using a series of spin-traps. Free Radical Biol. Med. 30, 1154-1162.

Turnbull, S., Tabner, B. J., El-Agnaf, O. M. A., Moore, S., Davies, Y. and Allsop, D. (2001b) a-Synuclein implicated in Parkinson's disease catalyses the formation of hydrogen peroxide in vitro. Free Radical Biol. Med. 30, 1163-1170.

Varadarajan, S., Yatin, S., Aksenova, M. and Butterfield, D. A. (2000) Review: Alzheimer's amyloid b-peptide-associated free radical oxidative stress and neurotoxicity. J. Struct. Biol. 130, 184-208.

Reply from the authors:

To the editors,

As Tabner et al. mention, the ESR experiments seem to have a number of potential problems. Therefore, it is not easy to interpret its results. We thus used the expression "Abeta-associated free radical generation " instead of Abeta-derived free radical generation to show our results. Our careful inspection did not show any definite spectra in the experiment using Abeta40 and Abeta42 peptides with deferroxamine or catalase. It was also the case with the experiment using PBS only. Our experiments using Abeta-(25-35) peptides showed the clear differences between the results obtained under the different experimental conditions described above. These results will be published in the near future (Monji et al. 2001). Anyway, I agree to the hypothesis that hydrogen peroxide plays an important role in the pathophysiology of Alzhemer's disease and transition metals such as Fe and Cu are involved in its generation with Abeta peptides while I think Abeta-derived peptidyl free radicals are also important considering our results in the experiments using ESR spectroscopy along with Th-T fluorometric assay and CD spectroscopy according to Kay's hypothesis of mechanochemical mechanism for peptidyl free radical generation by amyloid fibrils (Kay et al. 1997). Last but not least, I thank you for your interest in our study.

Akira Monji
Department of Neuropsychiatry,
Graduate School of Medical Sciences,
Kyushu University
Fukuoka, JAPAN

References

Monji, A., Utsumi, H., Ueda, T., Imoto, T., Yoshida, I., Hashioka, S., Tashiro, K.,Tashiro, N. (2001) Amyloid-beta-protein (Abeta) (25-35) - associated free radical generation is strongly influenced by the aggregational state of the peptides. Life Sci., in press.

Kay, C.J. (1997) Mechanochemical mechanism for peptidyl free radical generation by amyloid fibrils. FEBS Lett. 403, 230-235.