As upstream and early-development teams generate increasing numbers of samples, downstream purification must become faster, more flexible, and easier to scale. To demonstrate a practical and bench-saving solution, we configured three ÄKTA go™ chromatography systems to run in parallel from a single UNICORN™ software client connected to a shared database.
A combined workflow, affinity capture followed by inline buffer exchange/desalting, was implemented within a single method using scouting to run multiple samples, supporting multi‑step purification with minimal hands‑on time. Across twelve bispecific antibody runs (four per instrument), UV, conductivity, and pH profiles were highly consistent, with no observable drift in column pressure or peak asymmetry over the evaluated cycles.
Host cell protein (HCP) levels decreased from >100,000 ppm in the feed to <200 ppm after purification (>99% reduction), and SEC analysis showed 2–3% aggregates with approximately 1% low molecular weight forms.
These results demonstrate how parallel purification using the ÄKTA go™ system helps scientists increase throughput and streamline early-stage purification to address a common bottleneck in early-stage biologics development.
Integrated affinity capture and desalting workflow enhances efficiency
Early-stage programs increasingly require purification of many small-volume samples generated from parallel upstream experiments. To prevent purification from becoming a bottleneck for throughput, laboratories need approaches that increase sample capacity while minimizing operator intervention and bench space.
ÄKTA go™ is a compact chromatography system designed for flexible method development and small‑scale purification. Its modular configuration, compatibility with multiple column types, and UNICORN™ control software support multi‑step workflows in a range of purification schemes.
In this study, three ÄKTA go™ systems were operated in parallel controlled from a single UNICORN™ client to evaluate reproducibility and workflow efficiency for integrated affinity capture and desalting of antibody samples (Fig 1). This setup demonstrates early-stage multi‑clone, or multi‑construct purification during development, where many similar samples must be processed reproducibly with minimal operator time. The purification protocol was repeated four times on each system, resulting in 12 runs in total.
Fig 1. In this study, three ÄKTA go™ chromatography systems were operated in parallel controlled from a single UNICORN™ client to evaluate reproducibility and workflow efficiency for integrated affinity capture and desalting of antibody samples.
Materials and methods
Samples and columns
Two clarified, antibody harvest materials were used in this study: a monoclonal antibody (mAb8) for method definition and system qualification, and a bispecific antibody (bsAb2, titer at harvest 4.5-5.2 g/L) generated from parallel small‑scale bioreactor cultivations. A total of twelve bsAb2 purifications were performed, four on each system.
Affinity capture of mAb8 was performed using HiTrap™ MabSelect™ PrismA columns, packed with a high‑performance Protein A resin designed for robust monoclonal antibody purification. For the bispecific antibody runs, capture was carried out using HiTrap™ Mab Select™ VH3 columns (1 mL), containing an engineered ligand with specificity for light‑chain–containing antibody formats, enabling selective purification of the bispecific construct.
Following affinity capture, buffer exchange and desalting were performed using HiTrap™ Desalting columns, packed with a size‑exclusion resin optimized for rapid desalting and sample conditioning prior to analysis.
System configuration and software environment
Three ÄKTA go™ instruments were connected to a shared UNICORN™ 7.13 database and controlled from a single client.
Each system was equipped with:
- Extra inlet valves, V9-ImA and V9-ImB, with 6 buffer inlets each ideal for modern resins like MabSelect PrismA™, and Capto™ AVB
- Sample inlet valve, V9-ImS, with 5 samples ports plus a buffer inlet for sample chasing
- Column valve, V9-Cm, supporting up to three columns
- pH valve, V9-pH, for monitoring of pH
- Outlet valve, V9-O with 10 outlets to allow for flowthrough collection
- Fraction collector F9‑T in a tunnel under each ÄKTA go™ system to save bench space
- A 10 mL capillary loop was installed to accommodate intermediate elution peaks.
UNICORN™ 7 was installed using a centralized database and licensing server. Because each system requires a licensing server to operate, all three systems could be monitored and controlled through a single client interface yet still needed a dedicated PC to accommodate the server. Mini-PCs can be used for this purpose to save bench space.
In addition to standard evaluation tools, the workflow incorporated the Trending Tool 2.1, which compiles key parameters such as UV peak shape, column pressure, and conductivity into comparative plots for quick assessment of column health and run‑to‑run consistency. This supported rapid evaluation of reproducibility and helped identify any deviations during the multi‑step purifications.
Method design
The method was designed as a seamless workflow covering affinity capture, desalting, and affinity column regeneration (Table 1). It started with the equilibration and loading of the affinity column, followed by washing and elution steps that collected the target protein into a loop. After elution, the flow path was rinsed to remove any residual elution buffer before moving to the desalting step. With the desalting column equilibrated, the collected affinity eluate was applied and desalted. Once desalting was complete, the process transitioned back to the affinity column, which was then regenerated through strip, clean-in-place (CIP), and re-equilibration steps, restoring the column to its original state and completing the cycle.
This entire workflow is executed within a single, validated method, employing scouting functionality to enable multiple runs without altering the method structure. This ensures consistent operation, maintains traceability across affinity and desalting steps within one result, and allows controlled flow-path switching while avoiding unintended actions.
How to implement custom text instructions
- Select any phase (in this example, Column Wash was used for simplicity) and set all values to zero.
- Once the parameters are set, switch from Phase Properties to Text Instructions.
- Locate the relevant phase in the text instruction list—here, the Column Wash phase.
- Expand the phase by clicking the “+” icon.
- Select BaseSameAsMain—this is where custom text instructions can be inserted.
- From the Instruction Box, choose Flowpath, then Column valve, and finally select the desired column position.
This approach allows the flow path to be updated without losing access to the user interface for that phase, making it a practical solution for dynamic column switching during a run.
Other workflow configurations may also be suitable depending on sample stability and study requirements (see web article). While predefined methods are easy to set up, they generate separate results for affinity and desalting that affect data handling. In contrast, combining affinity and desalting into a single method with scouting provides an efficient setup: samples spend minimal time in acidic conditions, results are consolidated, save prompts are reduced, and users gain on‑the‑spot flexibility with only minor text guidance needed for flow‑path adjustments.
Table 1. Method structure with selections of inlet, %B, flowrate and phase length.
| Predefined phase | Phase | Inlet | %B | Flow rate (mL/min) | Length (mL) | Notes | |
| Affinity capture | Equilibration | Equilibration | A1/B1 | 0 | 5 | 15 | |
| Sample amplification | Sample application to PrismA | Sx/A1 | 0 | 2.5 | 15 | Used Scouting to vary the sample inlet position. Finalize sample injection with 25 mL A1 | |
| Equilibration | PrismA high salt wash | A2/B1 | 0 | 5 | 50 | ||
| Equilibration | PrismA low pH wash | A3/B1 | 0 | 5 | 15 | ||
| Elution | PrismA elution | A4/B1 | 100 | 4 | 15 | Peak to loop, start level 20 mAU, end 80 mAU, max peak volume 9.8 mL, wash inlet A4 (PBS) | |
| Equilibration | Rinse flow path in bypass | A4/B1 | 0 | 10 | 10 | 1 line text instruction: "Column valve: By-pass, Downflow" | |
| Desalting | Equilibration | Switch to column position 2 | n/a | n/a | n/a | n/a | Equilibration phase, 1 line text instruction: "Column valve: 2, Downflow" |
| Equilibration | Desalt equilibration | A4/B1 | 0 | 10 | 50 | ||
| Sample amplification | Sample application to Desalt | A4/B1 | 0 | 10 | 15 | Filling the loop using: Manual load, Loop type: capillary loop | |
| Column wash | Desalting | A4/B1 | 0 | 10 | 100 | Fractionation 96 well plates, 1.8 mL, peak fractionation start 80 mAU, end 50 mAU. | |
| Affinity regeneration | Equilibration | Switch back to column position 1 | n/a | n/a | n/a | n/a | Equilibration phase, 1 line text instruction: "Column valve: 1, Downflow" |
| Equilibration | PrismA strip | A1/B2 | 100 | 5 | 15 | ||
| Equilibration | PrismA CIP | A1/B3 | 100 | 1 | 15 | ||
| Equilibration | PrismA re-equilibration | A1/B1 | 0 | 4 | 20 |
To ensure straightforward operation across all three ÄKTA go™ chromatography systems, Method queues were used to start the workflows from a single command (Fig 2). This setup simplified operation by allowing all systems to be initiated together from one interface, reducing manual steps and making the workflow easier to manage.
Fig 2. Method queue interface enabling simultaneous workflow initiation across all connected systems for simplified operation.
Analytical methods
Elution fractions were analyzed for:
- HCP with Gyrolab immunoassay technology
- Aggregates by SEC
Results: Consistent chromatographic performance across multi-system operation
System performance and stability
All three systems operated in parallel as expected in this configuration. Pressure‑stability testing demonstrated that buffer containers could be placed on the floor or on a cart without inducing vacuum‑related disturbances at low flow rates (1–5 mL/min). A long NaOH CIP-step eliminated bubble formation observed during initial tests.
Fig 3. The chromatogram shows a representative run of the mAb8 test runs. Blue highlight Affinity capture segment, green highlight desalting, red highlight affinity column regeneration.
As shown in Figure 2, the flowthrough was successfully fractionated, the affinity‑elution peak was directed into the loop (with Out1 connected directly to the syringe port), and the desalted peak was fully processed, with all fractions collected as expected.
For subsequent runs, the UV, conductivity, and pH profiles from this run (ÄKTA go™ 1, S1 run) were selected as reference curves for evaluating the subsequent VH3 elution profiles (Fig 4). For users, any deviation in pH or conductivity during equilibration is a useful signal to pause the run and check all inlets, preventing sample loss or an improperly equilibrated column.
Fig 4. System control during a VH3 test run with reference curves (dashed lines). In this VH3 run, the equilibration buffer’s pH and conductivity match the previous run, while the feed shows slight variation.
Chromatographic reproducibility
Across 12 runs, UV peak profiles and conductivity traces were highly comparable, with only minor variation attributable to feed differences and loading during scouting. The data was analyzed using both Eval (Figure 4) and the Trending Tool 2.1 extension (Figure 5). The Trending Tool offers rapid visual feedback on column health, making it easy to spot changes over multiple cycles. Across the first runs, neither column pressure nor UV asymmetry showed any significant shifts. It is worth noting that the VH3 columns were initially run with bsAb2 feed to validate the reference curve before the sharper elution runs.
Peaks from the desalting segments can be integrated manually in Eval or processed in bulk using the Trending Tool. The resulting data can be exported to Excel for manual concentration calculations. Alternatively, extinction coefficients can be entered directly into Eval, allowing both calculated and user‑defined values to be stored within the software.
The following results highlight the consistency and performance of the parallel purification workflow across the 12 bsAb2 runs:
- UV, conductivity, and pH traces showed high reproducibility
- No trend in drift in column pressure or UV asymmetry was observed
- Trending Tool enabled rapid visualization of column performance across runs
Fig 5. Representative comparison of chromatograms for ÄKTA go™ 1 runs (Runs 1–4) using the Evaluation module. Overlay of the first four runs illustrate consistency in chromatographic performance.
Fig 6. Representative comparison of ÄKTA go™ 1 runs (Runs 1–4) using Trending Tool 2.1. Visualization of data shows consistency in chromatographic performance and highlighting any emerging trends across runs.
Analytical results
In this example workflow, HCP levels were reduced from high‑ppm feed concentrations to below 200 ppm after purification, corresponding to >99% reduction. SEC analysis confirmed acceptable aggregate levels with 2-3% aggregates and approximately 1% low molecular forms.
Discussion
The data shown here represents a typical example of expected chromatographic performance based on established platform knowledge and internal testing and is used to illustrate an automated workflow design and data handling approach. The results demonstrate that ÄKTA go™ can be effectively used in a parallel configuration to increase downstream throughput with a high level of automation. The ability to run three systems from a single UNICORN™ client simplifies workflow management, particularly during multi‑sample and multi-step purification workflows.
Method execution and time efficiency
With three ÄKTA go™ systems operated in parallel and up to five scouting runs per system, this setup enables up to twelve samples to be processed in a single unattended run.
Executing the workflow with the two-step purification in a linear progression, where each task is completed before the next begins is a traditional procedure that works fine when samples are limited. While straightforward, it often requires continuous supervision and extends the total timeline.
The model outlined in this article leverages parallelization and automation, allowing overlapping tasks and overnight processing. This approach significantly reduces hands-on time and accelerates the delivery of results.
By comparing these models over a standard workweek, Table 2 highlights how integrated workflows can optimize resource use, minimize delays, and improve throughput without compromising quality.
Table 2. A conceptual comparison of two approaches for executing a two‑step purification workflow over a standard workweek. In a traditional sequential setup, affinity capture, desalting, and column regeneration are performed one step at a time and are typically constrained by operator availability. The integrated workflow illustrates how coordinating these steps and enabling unattended operation could allow activities to overlap and make more efficient use of system time. The comparison is intended to illustrate potential differences in hands‑on effort and scheduling, rather than to represent a measured time study.
| Workflow aspect | Sequential workflow | Integrated workflow | Potential benefits |
| Execution of purification steps | Affinity capture, desalting, and regeneration performed one after another | Multiple steps coordinated within a single method or schedule | The integrated approach reduces waiting between steps |
| Operator involvement | Frequent manual setup, monitoring, and intervention | Reduced manual interaction once the workflow is initiated | Operators can concentrate on other tasks |
| Use of system time | Active runs interspersed with idle periods | More continuous use, including unattended operation | Systems can be used more efficiently |
| Scheduling flexibility | Runs typically limited to working hours | Runs can extend beyond working hours | Results are available earlier without additional staffing |
| Sample handling | Multiple start and stop points between steps | Fewer interruptions between purification stages | Reduced sample handling simplifies workflows and data review |
The modular hardware design supports multi‑step purification on a single compact system, enabling affinity capture, desalting, and column regeneration within one method. Scouting functionality further enhances versatility by allowing on‑the‑spot adjustments to sample volume and number of runs without modifying the validated method.
The small footprint of the ÄKTA go™ systems, combined with flexible buffer placement options, makes this setup well‑suited for laboratories with limited bench space. The reproducibility of chromatographic performance and the strong HCP reduction demonstrate that parallel operation does not compromise purification quality.
Conclusions: High purity outcomes enable reliable, scalable early-development purification
Parallel operation of three ÄKTA go™ systems from a single UNICORN™ client is feasible, reliable, and highly efficient. The configuration offers:
- High throughput: up to 15 samples per setup
- Versatility: multi‑step purification within a single method
- Time efficiency: reduced operator intervention and rapid evaluation
- Compact footprint: minimal space requirements
- Reproducible performance: consistent chromatographic behavior and strong impurity reduction
Overall, the results support the use of parallel ÄKTA go™ systems as a practical, high‑throughput platform for early-stage downstream development. The reproducibility, automation potential, and efficient use of space make this approach an attractive option for organizations seeking to accelerate candidate screening, method development, or small‑scale purification workflows without increasing staffing or equipment demands. The approach also benefits from built‑in parallelization and automation, allowing steps to run concurrently and continue overnight, which decreases manual effort and shortens overall timelines.
Get started: ÄKTA go system – free e-learning
Boost your research with automated parallel protein purification