Luminex Typing

The Luminex technology is based on the use of 5.6 micron polystyrene microspheres (beads) each internally dyed with a unique combination of red and infrared dye. The combination of different intensities of the two dyes allows for the identification of each bead by its unique signature when excited by a laser beam. This permits multiplexing of up to 100 reactions in a single tube. The surface chemistry of the beads allows them to be chemically coated with a number of different targets, including HLA sequences and antigens. The beads can therefore be used to interrogate samples for the presence or absence of specific analytes. The Luminex platform uses the principles of flowcytometry to stream beads in single file past a pair of lasers. A red laser is used to excite and therefore identify the specific bead and a green laser is used to excite and therefore identify any reporter dyes captured on the beads during the assay. As both the bead identification and reporter dye readings are made on each individual bead, a multiplex system can be developed with, typically, up to 100 beads. The probes used to coat the 100 beads are carefully selected such that individual analytes of interest can be identified by unique reaction patterns of the beads.

The application Luminex to HLA typing

HLA typing using Luminex is a reverse polymerase chain reaction sequence specific oligonucleotide (PCR-SSO) system which involves PCR amplification of targeted regions within the MHC class I or II regions with group specific primers, followed by a process of probing the amplicon with Luminex beads, each coated with sequence specific oligonucleotide probes to identify the presence or absence of specific alleles. The assignment of HLA type is then based on the reaction pattern observed, compared to patterns associated with published sequences.

Primers used for the amplification are biotinilated. Amplification can then be either symmetrical, which therefore requires a denaturation step to create single strands or can be asymmetrical to generate an excess of a single strand. The single stranded product is then hybridised with a multiplex of up to 100 beads, all of which can be uniquely identified by their internal dyes and all of which are selectively coated with specific oligonucleotide sequences. The amplified DNA hybridise to complementary DNA probed on the beads.  A washing stage may then be required depending on the Luminex typing kit used. Bound amplicon is detected by labelling with a Streptavidin – Phycoerytherin (SAPE) conjugate, with Streptavidin binding to the biotin used to label the primers and phycoerytherin serving as the reporter dye for the presence of bound amplicon. Again a wash step may be required depending on the kit in use.

The Luminex platform is used to identify any SAPE bound to the beads. The observed reaction patterns are used to assign HLA type. Positive and genitive control beads are used to quality control the typing test.

The current Luminex kits on the market tend to yield mostly medium resolution HLA types, though high resolution results are occasionally obtained. Suppliers are experimenting with high definition beads and increased numbers of beads in the multiplex in order to develop systems for high resolution HLA typing.

The advantages and disadvantages of Luminex for HLA typing

The Luminex methodology for HLA typing combines some of the speed typically associated with a PCR-SSP technique, with the high sample throughput of an SSO technique. This gives the ability to rapidly type a large number of samples with high reproducibility. The combinations of speed and reproducibility, together with the removal of the need to maintain and validate in-house methods, form the main advantages of Luminex as used for HLA typing. The technique is fairly robust and requires very little DNA. The methodology also makes better use of laboratory staff with fewer staff required to test the same number of samples when compared to SSP or other traditional SSO techniques.

The Luminex technique as used by most H&I laboratories is semi-automated but can be fully automated, contributing to the speed with which results can be obtained. There does however remain a need to experienced scientists to check, confirm and if required modify, the software proposed HLA types.

Another advantage of Luminex compared to PCR-SSP is the reduction in use of or even elimination of the use of agarose gel electrophoresis and its associated use of ethidium bromide and the H&S risks associated with that.

A potential disadvantage of the Luminex methodology is that even though it is rapid, it is still not as rapid as PCR-SSP and may therefore be unsuitable for use in an on call situation where a rapid turnaround of results is needed. In addition, the system may be better suited to batch testing of samples rather than the single sample testing typical of the on-call situation. In this situation the Luminex system may have a disadvantage both in terms of cost and speed. However, some laboratories do use it as a backup technique.

Another current disadvantage of the Luminex methodology is that results are low to medium resolution and therefore require further testing to obtain high resolution results where required. Luminex suppliers have developed sets of beads capable of yielding higher resolution results. One potential problem is the current limit of 100 beads used in a multiplex and ways of increasing this number are being examined.

A further limitation of the Luminex technology is that is does produce a small number of heterozygous ambiguities, though suppliers claim this is less than traditional SSO techniques. Where heterozygous ambiguities are identified, specific probes can be developed to help resolve these. This however points to another disadvantage of the Luminex methodology for HLA typing. With traditional in house SSO techniques or with SSP techniques, new probes could be rapidly added to help identify new alleles or resolve ambiguities. Use of Luminex does rely on the suppliers rapidly updating their kits.

Given the rapid turnaround of batched samples and the high associated high throughput, the advantages of the Luminex methodology do perhaps outweigh the disadvantages. The Luminex kits are however relatively expensive compared to in house techniques though staff time for these in house techniques needs to be taken into account.

Relevance of Luminex Typing

Molecular techniques have long replaced serological typing as a means of doing HLA typing, especially as more and more alleles are being discovered. Over time, there have been a number of molecular techniques for low to intermediate resolution HLA typing including Sequence Specific Primers (SSP), Restriction Fragment Length Polymorphism (RFLP), Single Strand Conformational Polymorphism (SSCP) and various SSO and reverse SSO techniques, with hybridisation on cards or strips. These techniques have varying degrees of automation and manual input required. Luminex HLA typing is a reverse SSO technique which involves PCR amplification of targeted regions within the MHC class I and/or II with group specific primers, followed by a process of probing the amplicon with Luminex beads, each coated with sequence specific oligonucleotide probes to identify the presence or absence of specific alleles. The assignment of HLA type is then based on the reaction pattern observed, compared to patterns associated with published sequences.

The Luminex HLA typing kits on the market today give a resolution of results that means that for solid organ, platelets and most disease association typing, no further testing is required to get the resolution of results required. Testing can be undertaken at all the relevant loci including HLA-A, B, C for platelets plus DRB1, DRB3, 4, 5, DQA, DQB1, DPA1 and DPB1 for solid organs. For stem cell services, Luminex typing gives a resolution of results that allows the potentially matched related donors to be identified and further high-resolution testing undertaken for suitable donors. Most laboratories are however switching to Next Generation Sequencing (NGS) which means that Luminex typing may be less relevant to stem cell transplantation in the future.

In terms of cost, the Luminex systems does compare well with other rapid high throughput systems. Luminex can be fully or semi-automated and allows for much easier batching of samples compared to some of the other techniques. This means that a large number of samples can be rapidly HLA typed in a single run. On the other hand, each run does take several hours (4-6hrs) which means that Luminex is not necessarily suitable for single sample HLA typing as would be required for deceased donor typing for instance where a turnaround time of 4 hours is required. It may potentially be suitable as a backup technique. Overall, the ability to rapidly type a large number of samples to a good level of resolution at a reasonable cost means that Luminex typing remains highly relevant to H&I service provision today.

qPCR

One qPCR system used for HLA typing is LinkSeq, supplied by Linkage Bioscience, the other being QType, supplied by CareDX. These are real-time allele specific PCR SSP methods. Both are set up on a plate like traditional SSP, however the samples are read on a real-time platform instead of gel electrophoresis.

LinkSeq uses a DNA intercalating dye SYBR-Green which fluoresces brightly when it intercalates with double stranded DNA. The real-time PCR platform takes fluorescent readings of each well on the plate at different temperatures. The double stranded DNA melts during each cycle, releasing the SYBR-Green and reducing the fluorescence. These changes in fluorescence over the temperature range generates a melt curve that can be used to detect the presence or absence of allele specific amplification. QType on the other works by incorporating a hydrolysis probe which binds ton the sequence somewhere in between the primer binding sites. The hydrolysis probe includes a fluorescent dye on one end and a quencher on the other. The close proximity of these prevents the dye from fluorescing when excited. During amplification, the hydrolysis probe is degraded and the dye and quencher are no longer in close enough proximity to prevent excitation which can be detected.

The RCPath college guidelines set out an 8-hour turnaround time from a donor becoming identified to the tissue typing being available for a matching run. As much of this time is taken up with the processes before the sample arrives in the H&I laboratory, a lab typically only has a four-hour window in which to complete a HLA type for a deceased donor. Traditionally, deceased donor typing has been performed by various PCR-SSP techniques. These have a four-hour turnaround time meaning that it was not always possible to obtain a result with the target time. qPCR has found a niche in decease donor typing. The process is mostly automated, does not require gel electrophoresis and result transfer can be automated so that no manual transcription is involved. It also has a 2-2.5hr turnaround time. This has advantages not just in speed and quality of results but also in the time of on call staff and is now highly relevant to H&I service provision.

SBT

Principles of Sequence Based Typing (SBT)

DNA sequencing techniques were first described in the 1970’s by Gilbert & Maxam using a chemical sequencing approach and by Sanger using a chain termination method. Both lead investigators were awarded the Nobel Prize for chemistry for their work. The Sanger sequencing method is the simpler of the two and is widely used for Sequence Based Typing (SBT).

The sequencing step in sequence based typing is preceded by locus specific PCR amplification to generate templates for the sequencing step. The sequencing step requires single stranded DNA templates, DNA primers, a DNA polymerising enzyme, deoxynucleotidephosphates (dNTPs) and fluorescently labelled di-deoxynucleotidephosphates (ddNTP). At a minimum, HLA class I requires sequencing at exons 2 and 3 and HLA class II requires sequencing at exon 2. Other exons are often also sequenced to help increase the resolution. The amplified PCR products are divided into four separate sequencing reactions (eight if sequencing in the forward and reverse directions), each containing all four dNTPs (dATP, dCTP, dGTP & dTTP) and the DNA polymerase. To each reaction, one of the ddNTPs is added. The ddNPT will terminate chain elongation once incorporated into the growing sequence as they lack the 3’-OH group required for the formation of a phosphodiester bond between themselves and an incoming nucleotide. The products of the sequencing reaction are analysed using automated high-throughput DNA sequence analyzers. Modern DNA sequencers use capillary electrophoresis for size separation, detection of dye fluorescence and data output as peak traces on a chromatogram.

Up until very recently, SBT has been critical to Haematopoietic Stem Cell transplantation where HLA typing at high resolution may be needed in the related setting and is critical in the unrelated setting. Most laboratories are now however introducing NGS which is reducing the reliance on SBT.

In stem cell transplantation, Petersdorf et. al., and others have shown that the risk of GvHD increases with increasing numbers of mismatches in the Graft versus Host direction (i.e. HLA genes present in the patient but absent in the donor). Similar studies have also shown that the risk of graft failure increases with increasing numbers of mismatches in the Host versus Graft direction (i.e. HLA genes present in the donor but absent in the patient).  SBT can be used to obtain high resolution second field level for HLA-A, B, C, DR and DQ and in some cases for HLA DP for such transplants.

Sequence based typing is also an option in some disease association and drug sensitivity cases where a high resolution HLA type is required. Examples include HLA-B*57:01 typing for Abacavir drug hypersensitivity in HIV positive patients and HLA-DRB1*06:02 typing in Narcolepsy cases.

Relevance of Sanger SBT

SBT is carried out using the Sanger sequencing method in which the sequencing step is preceded by locus specific PCR amplification to generate target templates. The sequencing step is carried out using labelled ddNTP’s which cause chain termination. The products of the sequencing reaction are analysed using automated high-throughput capillary electrophoresis DNA sequence analysers.

Up until very recently, SBT has been highly relevant to Haematopoietic Stem Cell transplantation though most laboratories are now switching to NGS. Whichever technique is used, high resolution HLA class I and II typing has been shown to be relevant in the related HSCT setting. High resolution HLA-A, B, C, and DR have been shown to be relevant in unrelated HSCT and in cord blood transplantation. In the UK HLA-DQ and HLA-DP high resolution typing and matching are often carried out.

SBT is also an option in some disease association and drug sensitivity cases where a high-resolution HLA type is required. Examples include HLA-B*57:01 typing for Abacavir drug hypersensitivity in HIV positive patients and HLA-DRB1*06:02 typing in Narcolepsy cases.

The long-term relevance of SBT is in question with the emergence of NGS. One of the disadvantages of SBT is that it often requires further exons to be amplified to resolve ambiguities. These are resolved at first pass with NGS. NGS is already making an important contribution to registry donor HLA typing. As newer, faster NGS techniques emerge, they are being applied to patient and family donor typing and are set to replace SBT.

NGS

Principles of NGS

There are multiple NGS systems but they all follow the same key stages of target generation, library prep, clonal amplification and sequencing on the platform. The GenDx kit on the Illumina MiSeq system is described here:

• Long range PCR for target generation
  • Starts with HLA locus specific amplification for target generation
  • Currently, the whole of class I but not all exons in all class II are amplified. There are plans to expand to all class II exons
  • The amplification can be verified on a gel
  • The amplicon concentration can be determined using a DNA quantification system such as a Qubit
• Library Prep
  • Two enzymes are used to generate fragments
  • The fragments are then end repaired to generate blunt ends, followed by dA tailing to generate fragments with an average size of 400bp
  • The dA tailing generates binding sites for the illumina adaptors
  • The next step is the addition of the adaptors
  • After adaptor ligation, the DNA is size selected (>400bp) and cleaned up using magnetic beads
  • The fragments + adaptors are then indexed with a pair of indexes, one on each side which allows amplicons from different HLA genes and different samples to be pooled
  • This early pooling is one of the advantages of the GenDx system
  • The indices contain sequences required for cluster formation
  • After indexing, a second size selection and DNA clean-up is required
  • After indexing the libraries can be pooled to generate a single tube with all amplified loci for all samples
  • The concentration of the pooled library is then determined to generate an optimum sample for the flowcell
  • The library quantification can be carried out on a qPCR platform
• Clonal Amplification
  • The library is then set to be sequenced on the MiSeq
  • The library is flowed over the flowcell and the indices anneal to complimentary targets on the flowcell
  • The fragments are amplified into clonal clusters
• Sequencing reaction
  • The clonally amplified clusters are then sequenced in a ‘sequencing by synthesis’ method
  • Just one base is added in each cycle of the ‘sequencing by synthesis’ reaction
  • The fluorescence of each incorporated base is captured in each cycle and the incorporated base recorded
  • Sequencing takes place in a ‘paired-end’ approach, with sequencing from each end, rather than a single direction
• Analysis and Reporting
  • NGS Engine is used for data analysis
  • No additional GSSP’s etc. required
  • Gives ‘unambiguous’ allele level HLA types

NGS has several advantages over traditional SBT.

• NGS allows multiple loci from multiple samples to be sequenced in parallel and is therefore allows for high throughput compared to SBT 
• SBT sequences single strands while NGS has a depth of read often greater than 30, resulting in fewer ambiguities
• The paired end sequencing approach of NGS makes it straight forward to phase-resolve
• NGS generates allele level HLA typing without requiring further amplification and sequencing as often required for SBT
• The cost of NGS has come down in recent years and is now comparable to, if not cheaper than SBT when other factors such as staff time are taken into account