Oiling the toxic wheels: Evidence for the role of haptoglobin phenotypes in toxic oil syndrome

Toxic oil syndrome revisited

It is now 30 years since toxic oil syndrome struck in Spain with devastating consequences. It is a food-borne disease that is estimated to have killed at least 2500 people with thousands more affected to this day with different degrees of disability.

There was a swathe of acute symptoms which developed into chronic conditions such as motor neuropathy, musculoskeletal pain, sclerodermia, pulmonary hypertension, liver impairment, hypothyroidism and diabetes mellitus.

And the sad point is that it could clearly have been avoided. Toxic oil syndrome was caused by the ingestion of a rapeseed oil that had been treated with aniline. It was initially intended for industrial use but had been released and sold by unlicensed street vendors as cooking oil. During its treatment at a refinery, a series of anilides and fatty acid esters of the aniline derivative 3-(N-phenylamino)-1,2-propanediol (PAP) were produced and they are thought to be responsible, although the jury is till out.

Not all members of the public who took the contaminated oil were affected and this led scientists to believe that there might be some personal factors that made individuals more susceptible to the disease. Preliminary work on gene expression comparisons between control subjects and those with toxic oil syndrome implicated certain phenotypes of the haptoglobin, a protein found in the blood.

Now, more evidence supporting their involvement has been unearthed by a team of researchers in Spain. Emilio Gelpi from the Barcelona Biomedical Research Institute CSIC and colleagues from the Institute of Advanced Chemistry of Catalonia, Barcelona, and the Institute of Health Carlos III, CISATER, Madrid, adopted a proteomics approach.

Serum secrets

The team analysed serum from patients displaying chronic symptoms and from people living and eating in the same home as patients but who were not affected themselves. The samples in each group were pooled according to age and gender in order to smooth out individual differences and enhance any traits associated with the disease.

Firstly, the sera samples were treated with immunodepletion kits to remove the abundant protein albumin, then the remaining proteins were extracted for two-dimensional gel electrophoretic separation. Image analysis following staining revealed a total of 329 protein spots, of which 35 were up- or down-regulated between the patient and control samples.

Among these, different phenotypes of haptoglobin were identified by mass spectrometry as being some of the more clearly differentially expressed proteins.

Haptoglobin is a tetramer consisting of two heavy and two light chains, denoted beta and alpha, respectively. The beta chains always have the same structure but the alpha chains appear in one of three forms, alpha-2, alpha-1s and alpha-1f. So, within the two alpha chains, an individual can have one of six possible combinations of light chains.

In this case, the gel spots comprised mixtures of the individual phenotypes, so the team chose a different approach to identify them, still based on mass spectrometry but excluding the prior gel electrophoresis step.

Haptoglobin phenotypes indicted

The sera were analysed following a preliminary reduction step to break the disulphide bonds between the chains in the tetramer. Albumin depletion was not employed due to the number of samples but a clear mass spectral peak was still observed for the alpha-2 chain.

However, the peaks for the alpha-1s and alpha-1f chains coalesced because the mass difference between them is just 0.043 Da and the mass spectrometer did not have the required resolution to separate them.

This problem was solved by treating the serum with O-methylisourea, a reagent that converts lysine amino acid residues into homoarginine. The alpha-1s chain contains eight lysine residues but the alpha-1f chain contains nine, due to a small change in the peptide sequence. After derivatisation, the mass spectral signals for the two chains differed by 42 Da, so they could be clearly identified and differentiated.

Having positively identified the haptoglobin phenotypes, the researchers could confirm that the alpha-1s chain was over-expressed in toxic oil syndrome patients and the alpha-1f chain was under-expressed.

When the proportions of patients and controls with a particular haptoglobin phenotype were compared, some of them appeared to correlate with the disease. Out of 27 individuals who had the alpha-2/alpha-1s phenotype, nearly 70% were patients. Conversely, only 20% of those with the alpha-2/alpha-1f phenotype were patients.

No patients had the alpha-1f-alpha-1f phenotype, implying that it could confer resistance to the disease. The alpha-2-alpha-2 phenotype showed no difference between controls and patients.

The possession of certain haptoglobin phenotypes has been associated with many inflammatory diseases, including autoimmune disorders. Toxic oil syndrome is related to the immune system, with some features similar to autoimmune disorders, so it might not be unexpected to uncover the potential role of haptoglobin.

Gelpi concluded that there is a strong possibility of a genetic influence, with the haptoglobin 1s allele of the alpha chain conferring susceptibility to toxic oil syndrome. The haptoglobin 1f allele is probably a protective factor.

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.