The present invention relates to a method of detection of endotoxin in polypeptide compositions. More specifically, this invention relates to " he measurement of endotoxin obtained by σuantitating the amount of a volatile derivatized component (beta-hydroxymyristic acid) of endotoxin present in a sample of any polypeptide. compositions. The present invention can be used for evaluat-inc _f endotoxin content of products v.'hich are intended for human and veterinary use as parenteral therapeutic agents in clinical medicine, including products produced by recombinant DNA technology.

Polypeptides produced by recombinant DNA technology have only recently become available ^' for use as therapeutic agents in clinical medicine. Polypeptides such as human insulin, human growth hormone, alpha-interferon and tissue plasminogen activator have all been approved by the Food and Drug Administration (FDA) for the treatment of human diseases. Several other interferons (such as alpha- interferon, beta-interferon, and subsets of gamma- interferon) , interleukin-2 and tumor necrosis factor are now

in early clinical stages of trial at centers throughout the world as treatments for cancer. Many polypeptide factors are expected to be manufactured and tested clinically over the next several years.

However, the central problem in marketing suc products for use in clinical medicine is the development of quality control procedures for insuring the product's identity, purity and biological potency as required by the FDA. One of the most critical quality control tests required by the FDA for all drugs is the measurement of endotoxin in the final formulation. This test is particularly important for drugs manufactured by recombinant ^' DNA technology in microorganisms. The standard test for endotoxin measurement has been the li ulus test, which is based on the ability of endotoxin to cause the lymph of the horseshoe crab to clot. Although this bioassay is relatively sensitive, its main drawbacks, lie in the fact that it is not sufficiently specific, it is a difficult test to standardize and it is not easily reproducible.

The present invention establishes a more efficient method of measuring endotoxin, by production and detection of a volatile derivatized component unique to endotoxin.

The present invention relates to A Method for detection of endotoxin in & polypeptide composition comprising the steps of: ^' hydrolysis of the composition containing said endotoxin, to form beta-hydroxymyristic acid; esterifying the resulting hydrolysate with an alkanol; reducing the carbalkoxy groups present in the hydrolysate to hydroxymethyl groups; separating the volatile alcohols from the reduced hydrolysate; and detecting beta-hydroxy- myristyl alcohol by mass spectrometry.

The invention further relates to a method for detection endotoxin in a polypeptide composition comprising the steps: hydrolysis of the polypeptide composition; esterification of the resulting hydrolysate with an alkanol; acylation of free amino groups present in the hydrolysate; reduction of the carbalkoxy groups in the hydrolysate to hydroxymethyl groups; extraction of beta-hydroxymyristyl alcohol from the" ^" reduced hydrolysate;

derivation of the extracted alcohol; separating the volatile ^d erivatized product from the hydrolysate; ^■ and detecting an _d quantitating the volatile derivatized alcohol of beta-hy _d roxy _¬ myristic acid by mass spectrometry.

The present invention contemplates a method of detection of the presence of endotoxin in a polypeptide composition. The term endotoxin as used in the specification and claims also includes gram-negative bacterial pyrogen and lipopolysaccharide (LPS).

The method of the present invention involves converting endotoxin by hydrolysis into beta-hydroxymyristic acid, which is subsequently converted into beta- hydroxy yristyl alcohol, which can be detected by mass spectrometry after separation from ether hydrolysis products.

The method of producing the volatile derivative of beta-hydrβxymyristic acid can be accomplished by the steps of hydrolyzing the biological system containing endotoxin; esterification of the resulting hydrolysate; conversion of carbalkoxy groups present in the hydrolysate by reduction to hydroxymethyl (-CH-OH) groups and separation of the volatile alcohols from the hydrolysate.

A preferred method of producing the volatile derivative of ^" be'ta-hydroxymyristic acid can be accomplished by the following steps: a) hydrolysis of' endotoxin in the polypeptide. composition into its--component parts; b) esterification of the resulting hydrolysate with an alkanol which forms, inter alia, an alkyl ester of beta-hydroxymyristic acid;

c) acylation of a ino groups present in the hydrolysate; d) reduction of the carbalkoxy groups of the esters formed in the hydrolysate to CH ₇ OH; e) selective extraction of the beta- hydroxymyristyl alcohol from the reduced hydrolysate; f) derivatizing the extracted alcohol of step (e), preferably by trimethylsilylation of the free alcohol groups.

The silylated product can then be separated using known means, preferably by gas chromatcgraphy, although column chromatcgraphy may also be used. The volatile derivatized component can then be identified and quantified by mass spectrometry techniques.

Reduction cf carbalkoxy groups to hydroxymethyl groups is preferably carried out by using a borohydride reducing reagent. In such reduction, a ino groups of the hydrolysate, if present, will complex with the boron of the reducing reagent.

It is usually preferred to hydrolyze the resulting reduction mixture to destroy the amine-boron bonds in order to solubilize the reduction mixture.

In the extraction step previously described, this step is selectively accomplished by the choice of suitable solvents such ^* as haiogenated hydrocarbons, e.g., methylene chloride, ethylene chloride, dichloroethane and the like.

While the invention, as described has been found to

.be particularly useful in connection with measuring the endotoxin content of protein products produced by recombinant

DNA technology, it is to be understood that the invention is not limited thereto but may be applied with equal facility to

any biological system, and to any and all pharmaceutical products manufactured for parenteral use, i.e. anti-cancer drugs, anti-hypertension drugs, .anti-lipodemia drugs, antidepressants, etc.

The following examples assist in further detailing the subject invention herein without in any way limiting same.

EXAMPLES

1. Hydrolysis of Endotoxin

Endotoxin was broken down into its components by incubation with 6N HC1 for 10 minutes at 120 ^o C.

2. Derivatization Chemistry

A. Esterification

The esterification reaction was performed by first preparing diazomethane and then reacting diazomethane with the oligopeptide mixture. a. Preparation of Diazomethane

99 g of N-methyl-N-nitroso-N-nitroguanidine (MNNG) were placed in the inner tube of a Pierce Microgenerator apparatus. Then 400 uL of H_.0 were added to the MNNG to form a suspension. Subsequently, 1500 uL of diethylether were dispensed in the bottom of the generator and the generator was placed in an ice bath. Using a 1 mL syringe fitted with a needle, 450 uL ^* of 5N NaOH were then carefully injected into the inner tube containing MNNG. The generator then remained in the ice bath for 45 minutes and this reaction gave an approximately 60% yield of diazomethane.

b. Reaction of Diazomethane with Endotoxin Components

The amounts of diazomethane required were estimated as follows: 99 mg of MNNG yielded approximately 400 μmoles of diazomethane in 1500 uL and the number of moles - of diazomethane required was approximately 100 times the number of moles of protein that was hydrolyzed.

Dry MeO ^' H to a final volume of 100 uL was pipetted into the PICOTAG chamber and vcrtexed. 40 uL of dry MeOH was then pipetted into each sample tube and vcrtexed. 400 uL of diazomethane solution was then pipetted into the PICOTAG chamber and vortexed. After calculating the amount of diazomethane required as described above, 10 times this amount was pipetted into each sample tube and then vortexed. Any remaining diazomethane solution was added to the PICOTAG chamber. The closed PICOTAG then stood at. room temperature for 4 hoμrs. The excess reagent was removed from the PICOTAG chamber. The reagents from the sample tubes were then- removed in the PICOTAG workstation under vacuum (30-60 mTorr) .

B. Acylation

Freshly distilled Methyltrifiuoroacεtate (MeTFA) was prepared and then 500 uL of MeTFA was mixed with 500 uL of dry MeOh. *2» uL of triethyla ine (TEA) was added to each sample tube and vortexed. The reagent was removed in the PICOTAG workstation under vacuum (30-50 Torr). 500 uL of MeTFA: eOH was pipetted into the PICOTAG chamber and vortexed. 2 uL of TEA were added to each sample tube and vortexed. 6 uL of MeTF :MeOK were added to each sample tube and vortexed. The PICOTAG chamber was then pressurized four

times with N ₂ in the PICOTAG workstation. The closed PICOTAG chamber then stood at room temperature in the dark overnight. The excess reagents were then removed from the PICOTAG chamber and the chamber was washed with MeOH. The reagents were removed from the sample tubes in the PICOTAG workstation under vacuum (30 mTorr).

C. Reduction

Boron trideuteride (1M) in tetrahydrcfuran was used as the reducing reagent in this reaction. 800 uL of Boron trideuteride (B _{2 g} ) in tεtrahydrofuran (THF) was pipetted into the PICOTAG chamber and vortexed. 100 uL of 1M E ___,Db,/THF was then ^' pipetted into each sample tube. The PICOTAG chamber was pressurized four times with N_ , and then heated in the PICOTAG workstation at 90°C for 30 minutes. The chamber was periodically removed and vcrtexed.

The excess reagent was removed from the chamber, which vas then washed with dry MeOH. The chamber was vortexed before the MeOH was removed. Dry MeOH (20 uL) was then slowly added to each sample tube to decompose the reducing reagent. The reagents were then removed from the sample tubes in the PICOTAG workstation under vacuum (200 mTorr). After reduction was completed, all a ine groups were compiexed to the Boron atom.

D. Hydrolysis of Boranes

This hydrolysis reaction was performed in order to destroy the amine-boron bonds. This was done by using a mixture of HCl and MeOH. The HCl/MeOH (IN) was prepared by diluting and mixing a vial (1 mL) of 3N HCl with 2 mL of dry MeOH. lmL of IN HCl/MeOH was then pipetted into the PICOTAG chamber and vortexed.

40 uL of IN HCl/MeOH was then added to each sample tube and vortexed. The PICOTAG chamber was heated in the PICOTAG workstation at 97°C for 20. minutes. The excess reagent was removed from the PICOTAG chamber. The reagents were removed from the sample tubes in the PICOTAG workstation under vacuum (100 mTorr).

The above steps were then repeated as follows: IN HCl/MeOH (1 mL) was added to the PICOTAG chamber and vortexed, IN HCl/MeOH (40 uL) was added to each sample tube and vortexed. The PICOTAG chamber was heated for 20 minutes at 97°C. The excess reagent was removed from the PICOTAG chamber. However, at this stage, the reagents from the sample tubes were removed under vacuum of 45 mTcrr.

E. Extraction

The derivatized components of endotoxin were extracted with mεthylene chloride (CH-C1-) in order to prevent unwanted by-products from interfering with the gas chromatography and mass spectrometry. 200 uL of CH^Cl- was added to each sample tube and each tube was vortexed individually for 3 minutes. The CH ₂ C1_ was then decanted into pre-weighed pyrex tubes and saved (Batch I). This step removed by-products which were soluble in organic solvents (and the reduced derivative of beta-hydroxymyristic acid from the aqueous phase which contains the polyamino alcohol derivatives) .

A saturated solution of potassium carbonate (K ₂ C0 ₃ ) was prepared and washed with CH ₂ C1 ₂ and then filtered. With the sample tubes in horizontal position, 10 uL of the K ₂ C0 ₃ solution was then added to the top of each sample tube one at a time. The sample was then vortexed as the drop descended

to the bottom of the tube. The sample tube was tilted so the drop moved to the top of the tube and the vortex procedure was then repeated. This procedure ensured that the sample and the K ₂ CO-. solution were properly mixed. 200 uL of CH ₂ Cl ₂ were then added tc each sample tube, which were then vortexed for 1 minute, and the phases were allowed to separate. The εupernatent was then carefully decanted into preweighed pyrex tubes (Batch II). The supernatεnt then contained the polyamino alcohols. The aqueous portion which remained behind contained the by-products that were soluble only in water.

The CH ₂ C1 ₂ was removed frcm the sample tubes (Batch I and Batch II) in the PICOTAG workstation under vacuum, which was 150 mTorr. The sample tubes were then weighed to calculate the amount of material present. (If the weight is too high, it is likely that one agueous phase containing the K ₂ C0 ₃ has contaminated the sample. If this occurs, wash the sample with 200 uL cf CH-,C1 ₂ , decant and remove the CH ₂ C1 ₂ from the sample tubes under vacuum [150 mTorr]).

F. Silylation

For each 500 ug of original sample, 4 uL of pyridine and 30 uL of TMSDEA ( πmethylsilyl diethylamine) were added to samples frcm Batch I and Batch II; each sample was then vortexed for 1 minute. The PICOTAG cnambεr was then heated at 56°C for 20 minutes. This reaction added a tri- methylsilyl group to the free alcohol groups. The sample tubes were capped and stored at -20°C until assay. 3. Separation

The components of endotoxin were separated on a gas chromatograph and introduced into the ion source of the mass spectrometer. The derivatized sample (0.5 uL) was injected into an Ultra-1 capillary column (0.33 urn film thickness, 200

urn internal diameter, 50 m length) via an on-column injector. The analysis was performed with a flow rate of 0.41 mL/minute under the following conditions: Initial temperature: 70°C,

Initial time: 0, Rate (C/Min) = 1.5, Final temperature =

310°C, Final time = 5 minutes.

Heated Zones:

Oven (Standby) .Ξ.etpoint: 70°C

Oven (Standby) Limit: 325°C

Injection Port B Setpoint: 70°C

Injection Port B Limit: 325°C

Ion Source: 350°C

Transfer Line Setpoint: 32C°C

Transfer Line Limit: 350°C

Under these ccnditicns, the derivatized alcohol of beta-hydroxymyristic acid eluted at 47.67 minutes (retention index of 1950). Ions with m/e 257 and 221 werε not detected in the "spectra of any other ccmpcund in the hydrolysate of endotoxin.