The development of gamma-linolenic acid (GLA) as a pharmaceutical product
Lapinskas P. (1999)
Presented at: Speciality Chemicals for the 21st Century (International Seminar), 16-17 September 1999, Valbonne, France.

Metabolic role of GLA
Commercial development of GLA
Agricultural research and development - Evening primrose
Agricultural research and development - Borage
Agricultural research and development - Biotechnology
Failure of pharmaceutical registration applications


In considering the possibilities for new pharmaceutical products derived from plants for the 21st century, it is perhaps useful to look first to the past to see what has been done, what has succeeded and what has failed, and what might be done differently. This paper will therefore examine the development of gamma-linolenic acid (GLA) as a case history which may provide some lessons for the future.

Up Top of page

Metabolic role of GLA

GLA is a long-chain polyunsaturated fatty acid, with 18 carbon atoms and three methylene-interrupted double bonds in the molecule (figure 1).

Molecular structure of GLA

Figure 1.    Structure of gamma-linolenic acid (GLA)

The first double-bond, counting from the methyl end, occurs after the sixth carbon This is expressed in the abbreviated name for GLA, 18:3 n-6. The concept of a family of unsaturated fatty acids, with the first double bond at a given position but with varying chain lengths and degrees of unsaturation, is important in metabolic studies, as these acids may be synthesised, one from another, in a chain. There are two such families in mammalian metabolism, namely the w-3 and w-6, and GLA is an important member of the w-6 metabolic chain, as shown in figure 2.

Omega-6 metabolic pathway

Figure 2.    Metabolic pathway of n-6 fatty acids.

GLA is not normally found in the diet but is synthesised from dietary linoleic acid (18:2 n-6) by a desaturation reaction catalysed by the enzyme D6-desaturase. GLA is in turn converted to dihomo-gamma linolenic acid (DGLA, or 20:3 n-6) by elongation and thence into a range of other fatty acids, prostaglandins and leukotrienes which have wide-ranging regulatory functions, especially in the mediation of inflammation. GLA therefore has an important function as a precursor in the synthesis of these molecules, and also in its own right as a structural component of cell membranes. It follows, that any disturbance in the availability of GLA may have consequences which could lead to impaired function and possibly to disease states (Horrobin 1990a).

Linoleic acid, the precursor for GLA, is commonly available in the diet, so deficiency of GLA through lack of substrate would be unusual. However, a deficiency of GLA might arise either if the rate of formation by the D6-desaturase enzyme was inadequate, or if there was excessive uptake by the following metabolic steps. The former situation may arise if the D6-desaturase enzyme capacity is reduced, or if it is inhibited. A reduced rate of GLA formation has been associated with a number of environmental factors, such as excessive alcohol consumption, ageing and certain nutritional deficits, and also with various disease states, such as diabetes, eczema, pre-menstrual syndrome and cancer. Excessive consumption of GLA may be caused by high rates of cell division (e.g. in cancer), in inflammation, and in anti-viral responses.

Full accounts of the metabolism and potential uses of GLA have been produced by Horrobin (1990b) and Huang and Mills (1996).

Up Top of page

Commercial development of GLA

Given the range of circumstances which can give rise to a functional deficiency of GLA and the potential adverse effects of such a deficiency, it was a natural step to consider providing an exogenous source to supplement the body's existing levels, both as a healthfood product, and as a pharmaceutical product for the treatment of particular disease states associated with low endogenous GLA levels. There were several proponents of this approach in the mid-1970's, but there was no interest from the mainstream pharmaceutical industry, so it was left to them to establish their own commercial companies if they wanted to commercialise their ideas. At least two of them did so - Dr John Williams founded Bio-Oils Research Ltd (Bio-Oils) and Dr David Horrobin founded Efamol Ltd (now Scotia Holdings plc).

Their first task was to find a suitable source of GLA. Direct chemical synthesis was found to be extremely difficult for, although the molecule is relatively simple, each double bond can exist in one of two forms (cis and trans) of which only the cis form is normally found in vivo. Direct chemical synthesis gives a roughly equal mixture of the two forms at each position, so that only 1/8 of the final yield is the correct form with all double bonds in the cis configuration. Since it is considered that the trans forms may not only be inactive but may actually block the active form, this was clearly unacceptable (Sinclair 1990, Dutta-Roy 1996).

Although GLA is not a common component in the diet, it does occur in the seed oils of range of plant families including the Onagraceae (evening primrose), Saxifragaceae (borage, bugloss) and Rubaceae (blackcurrant, redcurrant) (Phillips and Huang 1996). Of these, the evening primrose (Oenothera spp) was considered to be the most suitable, primarily because of the very simple oil profile (which minimised the possibilities of undesirable side effects) and its potential for agricultural production.

Neither Bio-Oils nor Efamol were able to raise the capital required to fund the full-scale clinical trials which were required before a pharmaceutical licence could be obtained. Evening primrose oil was however acceptable for sale as a healthfood supplement, so both companies launched products directly to the consumer, with the intention of funding clinical research from the profits. Efamol was the more successful and was able to carry out a series of trials, resulting in the licensing of two pharmaceutical products based on evening primrose oil for eczema (in 1988) and mastalgia (in 1992).

On the basis of its health-food business and its registered pharmaceutical products Efamol (now renamed Scotia Holdings) was able to raise substantial amounts of investment funding, particularly during the 'biotech boom' of the early 1990's, which it used to increase substantially its rate of research into other disease states which showed a disturbance of GLA metabolism. Particularly promising areas included cancer, the long-term complications of diabetes and the side effects of radiotherapy.

One problem in tackling these diseases was that substantial quantities of GLA would be required to achieve beneficial effects, and that evening primrose oil, with just 8 - 10% GLA was too dilute, resulting in unrealistically high volumes of oil (20 or more 500 mg capsules per day) being required. Other plant sources of GLA offered higher concentrations: blackcurrant (Ribes nigrum L) has 17 - 19% and borage (Borago officinalis L) has 20 - 24%. However there were concerns that these oils were not functionally equivalent to evening primrose oil because of other components present (although this has been disputed) (Horrobin 1994, Bard et al. 1997) , so a concentration process was devised which allowed the production of high purity GLA (up to 99% pure), allowing up to a ten-fold reduction in the volume of oil for a given dose of GLA. The concentration process meant that, since any deleterious material would be left behind, the most cost-effective source of GLA could be used. After consideration of all the options, borage oil was chosen.

Up Top of page

Agricultural research and development - Evening primrose

At the time when interest in GLA began, evening primrose was just a wild plant, apart from some minor use as a garden flower, although it had been extensively studied for many years because of its unusual genetic properties (Harte 1994). These ornamental strains were used for the initial production of commercial crops, but they were difficult to grow on an agricultural scale. Research was started therefore in the mid 1970's to find the optimum methods for producing evening primrose seed crops, and to overcome some of the problems through plant breeding and selection.

There have been a number of such research programmes over the years, but the one with which I have been associated was the Efamol/Scotia programme, where I started as a plant breeder in 1978 and which I ran until its closure this year. In this programme, we aimed to solve the following problems:

Seed dormancy: Wild evening primrose can have substantial dormancy in the seed and can survive up to 20 years in soil before germinating.

Length of life cycle: Wild evening primrose is normally biennial whereas an annual form would be much more economic to produce

Weed control: evening primrose in the wild is a primary coloniser, but it is not a strong competitor and can easily become crowded out. This is particularly significant for the biennial forms, where it can be 9 months or more before a crop canopy is established.

Disease control: Septoria oenotherae can be a devastating disease in over-wintered crops and Botrytis spp can be serious at flowering time, especially in high rainfall areas or with heavy overhead irrigation.

Pest control: Pests have not been a major problem, although some local pests can be damaging, such as the alder flea beetle (Altica lythri) in the UK.

Bolting: In its biennial form, the evening primrose plant relies on the change in daylength from winter to summer to trigger the bolting response, and hence flowering. This sensitivity is mediated by a number of factors, including plant size, and there is a genetic component. Inadequate sensitivity (e.g. in a low-latitude area) can result in bolting failure.

Seed shedding: In wild evening primrose races, the capsules split as they ripen, thus shedding their seed. Since the capsules ripen over a period of weeks or months, there is substantial seed loss in a cropping situation.

In addition to these, there was a need to optimise the production methods at all stages of the crops production, from sowing through to storage of the harvested seed, so considerable work was needed in testing a variety of agronomic parameters, including sowing rates, row spacing, fertiliser rates, herbicide regimes, etc.

Very substantial progress was made in all of these areas, such that it may now be said that evening primrose has been domesticated as an agricultural crop. Production of improved, non-shedding varieties under mechanised agriculture in the UK, Holland, New Zealand and the USA has become a routine operation. In addition, the GLA content of the oil in the latest varieties is now normally over 12% and often over 15%, which is a substantial improvement over the original lines which yielded only 8 - 9%. (Lapinskas 1993, Fieldsend 1996).

Up Top of page

Agricultural research and development - Borage

With the realisation in the late 1980's that a concentrated GLA product would be need for advanced pharmaceutical products, and with the discovery of a cost-effective concentration process, it was clear that a similar improvement programme would be needed for borage. Some work had already been done with this species in 1982 - 1984, so the basic parameters and problems were known. The main issue was that, as in evening primrose, the seed ripened over a period of time, and was shed as it matured, although the physiology is rather different. A breeding programme was therefore established with the principal aim of overcoming this problem. It was found to be a more complex problem than in evening primrose, because the species is predominantly out-breeding, and the shedding character was more complex, being controlled by at least two genes. Again, however, the programme was successful, and stable, non-shedding varieties have been produced which is the critical step in the domestication of this species (Fieldsend 1995).

Up Top of page

Agricultural research and development - Biotechnology

When the agricultural research described above began, it was the only realistic route to the production of reliable and cost-effective supplies of GLA. By the 1990's however, there were alternatives available. Reddy and Thomas (1996) were the first to demonstrate that the D6-desaturase gene from an alien source (the cyanobacterium Synechocystis) could be transferred into a plant species (tobacco) to enable it to produce GLA, although the level of expression was very low. A substantially higher level of expression was achieved by transferring the D6-desaturase gene from borage into tobacco by Sayanova et al. (1997), giving a GLA content of 13.2% of leaf lipids, thus paving the way for the production of GLA in a mainstream crop such as oilseed rape at a small fraction of the cost of sourcing from either evening primrose or borage.

Up Top of page

Failure of pharmaceutical registration applications

Unfortunately, the early promise of the major new disease indications in the pharmaceutical sector was not realised in the form of successful licence applications. The application for the long-term complications of diabetes (diabetic neuropathy) was rejected by the UK authorities after an appeal in December 1997; the application for the amelioration of the side effects of radiotherapy was withdrawn after discussions with the European authorities shortly afterwards. This coincided with a radical change in the management team at Scotia Holdings, who took the decision to husband their resources and concentrate on a small number of key projects, none of which were based on GLA. Thus, while some work in this field does continue at a low level, and the original pharmaceutical products for eczema and mastalgia continue to be manufactured and sold, there is unlikely now to be any significant development in pharmaceutical uses of GLA from the company for many years (Anon 1998, 1999).

As part of the rationalisation, and in view of the dramatically reduced demand forecasts for GLA raw materials, Scotia management have closed their plant research and seed production facilities, and the staff have all been made redundant. The collaboration on biotechnology research is also being wound down. No further plant research is planned and any requirements for GLA-based new products for the future will probably be supplied from growers who already supply the healthfood market.

Up Top of page


The reader will be able to draw his or her own conclusions from the experience related above, but I would suggest that the following could be considered:

There are very substantial barriers to the development of new pharmaceutical products based on crude botanicals (as opposed to pure chemical compounds, even if plant derived). It is difficult to obtain patent protection for such products, and any other form of lesser protection is likely to be inadequate. There are therefore likely to be few such new products in future.

Where new pharmaceutical products are being developed by a young, scientifically-orientated ('biotech') company, there is a tendency to underestimate the regulatory hurdles which have to be cleared. It came as a surprise for instance, to Scotia, to be asked for animal toxicity data on evening primrose oil, which had already been taken by hundreds of thousands of humans for many years as a health food without ill effect.

The process of domesticating a new species, in order to obtain a particular chemical compound, is long, expensive and often only partially successful. Now that the technology exists to identify and transfer the appropriate gene to a species which is already domesticated, it is becoming more and more unlikely that any company would in future choose the old route for the development of a pharmaceutical raw material, unless there were truly exceptional circumstances.

On the other hand, with the substantial growth in sales of botanical medicinal products direct to consumers in recent years, and their hostility to genetically modified organisms (GMO's), it is quite possible that this may prove a more receptive market, with much lower barriers to entry, for those interested in developing new medicinal (if not pharmaceutical) products from plants. Opinion is likely to be divided, however, on whether this a good thing.

Up Top of page


I would like to thank Scotia Pharmaceuticals Ltd for permission to present this paper, and acknowledge the work of the many colleagues at the Scotia Plant Technology Centre who contributed to the domestication of evening primrose and borage, in particular Andrew Fieldsend, Nuala O'Connell, Mike Simpson and Ron Stobart.

Up Top of page


Anon, 1998.
Scotia Holdings PLC annual report and accounts 1997.
Scotia Holdings PLC, London, 48pp.

Anon, 1999.
Scotia Holdings PLC annual report and accounts 1998.
Scotia Holdings PLC, London, 48pp.

Bard, J.M., Luc, G., Jude, B., Bordet, J.C., Lacroix, B., Bonte, J.P., Parra, H.J., Duriez, P., 1997.
A therapeutic dosage (3g/day) of borage oil supplementation has no effect on platelet aggregation in healthy volunteers.
Fundam. Clin. Pharmacol. 11, 143-144.

Dutta-Roy, A.K., Campbell, F.M., Taffesse, S. and Gordon, M.J., 1996.
Transport of long-chain fatty acids across the human placenta: role of fatty acid binding proteins.
In: Huang, Y.-S. and Mills, D. E. (Eds.), 1996. Gamma-linolenic acid metabolism and its roles in nutrition and medicine. AOCS Press, Champagne, Illinois, 42 - 53.

Fieldsend, A., 1995.
Borage - a crop with a future?
The Biologist 42(5), 203-7.

Fieldsend, A., 1996.
Evening Primrose - from garden flower to oilseed crop.
The Horticulturalist 5(3), 2-5.

Harte, C., 1994.
Springer-Verlag, Berlin, 261 pp.

Horrobin, D. F., 1990a.
Gamma linolenic acid.
Rev. Contemp. Pharmacother. 1, 1-45

Horrobin, D. F. (Ed.), 1990b.
Omega-6 essential fatty acids: pathophysiology and roles in clinical medicine.
Alan R. Liss, New York, 595 pp.

Horrobin, D. F., 1994
Natural ¹ safe
Pharm. Tech. Europe 58, 14-15

Huang, Y.-S. and Mills, D. E. (Eds.), 1996.
Gamma-linolenic acid metabolism and its roles in nutrition and medicine.
AOCS Press, Champagne, Illinois, 319 pp.

Lapinskas, P., 1993.
Oil crops for the pharmaceutical industry.
In: Shewry, P.R. and Stobart, K. Seed Storage Compounds: biosynthesis, interactions and manipulation. Proc. Phytochem. Soc. Eur. 35, 332-42
(Full text)

Phillips, J.C. and Huang, Y.-S., 1996.
Natural sources and biosynthesis of g-linolenic acid: an overview.
In: Huang, Y.-S. and Mills, D. E. (Eds.), 1996. Gamma-linolenic acid metabolism and its roles in nutrition and medicine. AOCS Press, Champagne, Illinois, 1 - 13.

Reddy, A.S., Thomas, T.L. 1996.
Expression of a cyanobacterial D6-desaturase gene results in gamma-linolenic acid production in transgenic plants.
Nature Biotech. 14, 638-642.

Sayanova, O., Smith, M., Lapinskas, P., Stobart, K., Dobson, G., Christie, W.W., Shewry, P.R., Napier, J.A. (1997).
Expression of a borage desaturase cDNA containing an N-terminal cytochrome b5 domain results in the accumulation of high levels of D6-desaturated fatty acids in transgenic tobacco.
Proc. Natl. Acad. Sci. USA, 94(8), 4211-6.
(Abstract)    (Full text - external link)

Sinclair, H.M., 1990.
History of essential fatty acids.
In: Horrobin, D. F. (Ed.), 1990b. Omega-6 essential fatty acids: pathophysiology and roles in clinical medicine. Alan R. Liss, New York, 1 - 20.

Up Top of page

© Peter Lapinskas 1999-2012 Email Peter Lapinskas Last updated: 3 July 2012

Home      Services      Background      Publications      Resources      Contact