Loading the chip: frontfilling, backfilling and DDI

A mologram is two sets of lines on one surface, ridges and grooves. Loading a chip is two separate immobilization steps: the active ligand goes onto the ridges, and a matched, non-binding probe goes into the grooves. This page explains what each step does to the signal and how DNA-directed immobilization (DDI) puts each ligand in its place. It also shows how to set the ligand density.

Two steps on one surface

CMD reports the coherent mass difference between ridges and grooves. Each loading step therefore has a known effect on the immobilization trace.

Frontfilling
The active ligand binds the capture chemistry on the ridges. Mass goes onto the ridges only, so the immobilization CMD goes up.
Backfilling
A non-binding backfill probe binds the capture chemistry in the grooves. The grooves catch up with the ridges, so the immobilization CMD goes back down toward balance. The probe is chemically similar to the ligand but cannot bind the analyte.

The two regions carry orthogonal capture chemistries, so each step fills one region and leaves the other alone:

MD is the refractometric readout. It sees the average over both regions, so it keeps rising with every step that adds mass. A low immobilization CMD after backfilling does not mean the ligand is gone; MD shows that the mass is still there.

Backfilling is not a generic block, such as BSA or Tween in the running buffer, and it is not a coating. It is a separate, addressed immobilization that rebuilds the groove to mirror the ridge.

Load a chip, then add serum

Press Play to run four steps: frontfill, backfill, a blank of the sample matrix, then analyte in the same matrix. Set the backfill level and watch the immobilization CMD (violet) and MD (gray). Then switch to a chip with PEG grooves, which cannot be backfilled. In this model, CMD is the magnitude of the ridge-groove difference.

Model assumption: adsorption mismatch is set proportional to loading mismatch. Equal loading therefore matches this model, but does not guarantee a match in an assay. These are illustrative units, not tuning settings.

Backfilling suppresses coherent NSB; it does not stop fouling

An affinity-matched chip does not stay clean in serum. Matrix proteins still adsorb across the whole surface and can build up substantially in undiluted plasma. Backfilling does not stop that adsorption. It makes the adsorption similar enough on ridges and grooves that the coherent part stays small.

In the model, MD rises while the CMD baseline stays flat when adsorption is matched. A similar divergence in measured data is consistent with common adsorption being rejected; real MD traces also depend on bulk index and other disturbances. The fouling is real, so the chip and the fluidics still need cleaning and regeneration; the documentation describes how.

Three things can be matched, and they are not the same:

  1. Equal immobilization levels. The two regions carry comparable amounts of probe.
  2. Matched surface chemistry. The probe presents a similar face to the sample.
  3. Matched non-specific adsorption. Matrix proteins adsorb equally on both regions.

Only the third suppresses the background. The first two make it likely, not certain. A blank of the sample matrix tests the third directly. A reproducibly small CMD response to the relevant blank supports adequate matching under those conditions. It does not prove that no adsorption occurred or that the ligand remains active. A blank response warrants investigation alongside MD and repeats.

When backfilling matters

Coherent NSB grows with three factors: the affinity mismatch between ridges and grooves, the concentration of the background binder, and its affinity for the surface. It is therefore largest in crude, protein-rich samples such as serum, plasma and cell lysate, where abundant proteins are present at micromolar to millimolar concentrations. For these samples backfilling is an essential product step, and the chip needs reactive grooves.

In clean buffer or a purified protein in a defined buffer, uniform adsorption is already small in CMD. Chips with PEG grooves, [Oligo|PEG] and [Me-Tz|PEG], then do the job without a backfill step. The chips page lists the four surfaces.

What makes a good backfill probe

A backfill probe presents the same adsorption profile to the matrix as the ridge ligand, and never engages the analyte. In practice it is:

Typical probes:

Setting the backfill level

Comparable ridge and groove loading with chemically similar probes is a starting strategy. Equal immobilized mass, optical balance and matched non-specific adsorption are different properties. Neither saturation nor equal loading guarantees the adsorption match.

Changing the backfill level changes optical contrast. Excess groove mass can produce a zero crossing, but optical balance alone cannot establish matrix compatibility.

Judge the match from reproducible blank-matrix behavior under the intended assay conditions, alongside MD and appropriate controls. Use the immobilization trace as a guide. The documentation contains the tuning procedure; validate its starting strategy for your ligand and matrix.

DNA-directed immobilization

DNA chips carry 20-mer single-stranded capture oligos on the ridges. The ligand comes in with the complementary strand attached and hybridizes to its capture strand. The chip surface never changes: the ligand chemistry lives in the conjugate or the adapter, not on the chip.

Addresses
Different molograms can carry different capture sequences. Each sequence is an address: a conjugate goes only where its complementary sequence is. A chip carries up to 64 addresses, numbered 01 to 64, and sequence 00 is reserved for the backfill. With more addresses, more ligands share the 64 molograms and each gets fewer replicates. This is the multiplexing.
Strip and reload
Dehybridizing the duplex removes the whole ligand layer, ligand and bound analyte together. A fresh ligand is then loaded for the next cycle. No condition has to be found that strips the analyte but keeps the ligand alive.
Adapters and kits
Most ligands reach the chip through a DNA duplex. Conjugation kits attach the oligo to amines or thiols, an on-chip adapter couples amines, and splint ligation uses a DEL hit's own DNA tag. Oligo adapters also capture tags, biotin or antibodies. Each stays regenerable as long as the duplex is the weakest link. See adapters.

Two limits apply. Click chips bind the ligand covalently, with no duplex in between, so they cannot be stripped and are single-use per ligand. Some adapters release their ligand during the run, such as a His-tag on trisNTA. On a multiplexed chip, the released ligand can then rebind a neighboring address and contaminate it. Multiplexed chips need a stable attachment.

Favalli et al. (2026) use DDI to place DNA-conjugated library hits on their own molograms.

The chips page shows DDI as an animation. A pool of conjugates sorts itself onto its addresses, then the chip is stripped for the next pool.

Setting the ligand density

Density is a design choice. A dense surface gives a large signal, but analyte can then be captured faster than flow delivers it. Association then follows transport, and released analyte rebinds before it escapes, which slows the apparent dissociation. Choose the lowest density that still gives a reproducible response; the sensorgram page shows what transport does to a fitted curve.

On a DDI chip, the mixing ratio is a cleaner control than the loading time. Mix the ligand conjugate with the same complementary strand without a ligand. Every capture strand on the ridge still hybridizes, so the surface chemistry and the match to the groove stay the same. Only the fraction of active sites falls. Loading for a shorter time also gives fewer ligands. But it leaves single-stranded capture strands behind, and the ridge surface then depends on the loading time.

On click chips, the equivalent is to mix active and non-binding variants in a known molar ratio. In each case, check the result against the measured response, not the nominal ratio.

Dilute the ligand on the ridge

Set the fraction of active conjugate in the loading mix. The ridge below shows what lands. The plot shows a model binding curve at that density (violet) and the same interaction without a transport limit (dashed). Switch to a shorter loading to see what changes on the ridge.

The procedures are in the documentation.

The MACS® Matchmaker documentation gives concentrations, flow rates, contact times and stop conditions for each chemistry. See ligand immobilization, backfilling and NSB suppression, chips and surface functionalization and sensor regeneration.

After backfilling, the immobilization CMD is back near its starting value. Has the ligand come off the ridges?

Answer

No. CMD reports the difference between ridges and grooves, and backfilling filled the grooves to match the ridges. MD, which sees the average of both regions, still shows the full immobilized mass. A balanced CMD after backfilling is the goal; the blank matrix then shows whether the adsorption is matched too.